{
  "schema_version": 1,
  "corpus": {
    "label": "Newest 60 playable films in the September 5, 2026 audited inventory",
    "selected_at": "2026-09-17T18:15:56.984666+00:00",
    "inventory_cutoff": "2026-09-05T17:13:40.602033+00:00",
    "selection_rule": "All 621 in-window watch-playable records, ordered by timezone-aware publication instant descending, then video ID descending. Take 60. No genre or runtime exclusion.",
    "date_basis": "Archived official watch-response playerMicroformatRenderer.publishDate, not relative labels or titles.",
    "publication_range": [
      "2026-07-23",
      "2026-09-04"
    ],
    "eligible_count": 621,
    "inventory_sha256": "c2d0c8b530a824aaf5607dfcf4a2c6f82b2ede71094414f36a7f121edd6d87be",
    "limitations": [
      "Audited membership and current media acquisition are different snapshots.",
      "Historical local-source records do not establish that deleted media still exists.",
      "Shorts and Live are included through the audited deduplicated inventory."
    ],
    "not_claimed": "Not the latest 60 uploads on the channel on September 17, 2026."
  },
  "films": [
    {
      "id": "B-jjnydci50",
      "rank": 1,
      "title": "GPT-6 Astra with Ben Davis",
      "source_url": "https://www.youtube.com/watch?v=B-jjnydci50",
      "publication_date": "2026-09-04",
      "publication_timestamp": "2026-09-04T17:11:56-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "An interviewer asks what changed after difficult puzzles resisted people. Examples and an agent dashboard support the guest's enthusiasm, while acknowledged hints qualify his success claims.",
      "family": "interview",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and complete unverified ASR; analytical windows, not cuts. No playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 30,
          "role": "hook",
          "summary": "Surprise prompts a question about testing.",
          "visual": "Two seated men share a laptop.",
          "evidence_time": 15.5,
          "mechanics": "Two-shots alternate with the curly-haired guest's close-up.",
          "asr_anchor": {
            "word_index": 120,
            "source_time": 15.2
          }
        },
        {
          "start": 30,
          "end": 55,
          "role": "proof",
          "summary": "Cube-puzzle success includes an acknowledged hint.",
          "visual": "Twelve colored cubes precede an answer page.",
          "evidence_time": 40.5,
          "mechanics": "Laptop view precedes screen capture with guest inset.",
          "asr_anchor": {
            "word_index": 250,
            "source_time": 35.18
          }
        },
        {
          "start": 55,
          "end": 80,
          "role": "qualification",
          "summary": "A second puzzle also required assistance.",
          "visual": "Guest close-ups surround a seated two-shot.",
          "evidence_time": 65.5,
          "mechanics": "Faces replace the earlier evidence pages.",
          "asr_anchor": {
            "word_index": 430,
            "source_time": 57.2
          }
        },
        {
          "start": 80,
          "end": 105,
          "role": "discussion",
          "summary": "Guest explains parallel investigation and misdirection.",
          "visual": "Research dashboard lists branches and activity.",
          "evidence_time": 85.5,
          "mechanics": "Wide and enlarged dashboard samples retain guest inset.",
          "asr_anchor": {
            "word_index": 604,
            "source_time": 80.16
          }
        },
        {
          "start": 105,
          "end": 135.814,
          "role": "close",
          "summary": "Guest encourages larger collaborative-agent experiments.",
          "visual": "Faces, dashboard, two-shot, then publisher card.",
          "evidence_time": 130.5,
          "mechanics": "Conversation returns after the dashboard insert.",
          "asr_anchor": {
            "word_index": 822,
            "source_time": 108.04
          }
        }
      ],
      "editing_lesson": {
        "start": 30,
        "end": 55,
        "observation": "Laptop framing gives way to enlarged cubes and an answer page with guest inset.",
        "inference": "Interpretation: enlargement makes the anecdote inspectable without losing its teller.",
        "adaptation": "Show an original puzzle, enlarge its evidence, then reveal results with assistance disclosed.",
        "caution": "An answer page cannot establish repeatability or world-first status.",
        "evidence_times": [
          35.5,
          40.5,
          50.5
        ]
      },
      "codes": [
        {
          "code": "question-led",
          "beat_indices": [
            0
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            2,
            4
          ]
        }
      ],
      "uncertainty": "Dress photographs appear briefly within the answer-page insert, not as a dedicated demonstration; puzzle achievements remain the guest's claims.",
      "media": {
        "sha256": "0116e28b5179365eca7c09540d58be577b8be054208049a4184b241a6fa0a05b",
        "duration": 135.814,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:15:59.434162+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 490,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "0116e28b5179365eca7c09540d58be577b8be054208049a4184b241a6fa0a05b"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "bOC3DisEOfg",
      "rank": 2,
      "title": "Introducing GPT-6 Astra for developers",
      "source_url": "https://www.youtube.com/watch?v=bOC3DisEOfg",
      "publication_date": "2026-09-04",
      "publication_timestamp": "2026-09-04T13:22:31-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A presenter proposes delegating larger tasks with less supervision. Painting, output comparisons and a revised running task illustrate that proposition before an invitation to test it personally.",
      "family": "product-film",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and complete unverified ASR; analytical windows, not cuts. No playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 55,
          "role": "context",
          "summary": "Presenter frames broader delegation and computer operation.",
          "visual": "Desk presenter; capability heading overlays wider framing.",
          "evidence_time": 35.5,
          "mechanics": "Close-ups and desk views precede demonstration screens.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 4.7
          }
        },
        {
          "start": 55,
          "end": 90,
          "role": "action",
          "summary": "A photograph becomes a painting demonstration.",
          "visual": "Attached photo, painted laptop scene, presenter return.",
          "evidence_time": 70.5,
          "mechanics": "Input screen precedes Krita; speed disclosure appears later.",
          "asr_anchor": {
            "word_index": 312,
            "source_time": 56.38
          }
        },
        {
          "start": 90,
          "end": 136,
          "role": "proof",
          "summary": "Presenter claims richer creative outputs.",
          "visual": "Website comparison, garden rendering, then cityscape.",
          "evidence_time": 115.5,
          "mechanics": "Different artifacts interrupt recurring desk shots.",
          "asr_anchor": {
            "word_index": 572,
            "source_time": 94.06
          }
        },
        {
          "start": 136,
          "end": 187,
          "role": "action",
          "summary": "Running work receives a revised instruction.",
          "visual": "Code panels, task messages, rendered garden.",
          "evidence_time": 175.5,
          "mechanics": "Split-screen explanation precedes dark task UI and output.",
          "asr_anchor": {
            "word_index": 842,
            "source_time": 142.58
          }
        },
        {
          "start": 187,
          "end": 203.274,
          "role": "close",
          "summary": "Presenter invites testing on difficult tasks.",
          "visual": "Desk presenter followed by publisher card.",
          "evidence_time": 195.5,
          "mechanics": "Human framing replaces the technical demonstration.",
          "asr_anchor": {
            "word_index": 1150,
            "source_time": 190.36
          }
        }
      ],
      "editing_lesson": {
        "start": 55,
        "end": 90,
        "observation": "Photo attachment, recognizable painted desk and presenter return form three sampled states.",
        "inference": "Interpretation: a shared subject makes the transformation legible.",
        "adaptation": "Capture your workspace, show its submitted photo, reveal the resulting illustration, then assess it.",
        "caution": "The film labels playback 28-times faster; this is not latency evidence.",
        "evidence_times": [
          60.5,
          70.5,
          85.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            3
          ]
        },
        {
          "code": "output-montage",
          "beat_indices": [
            2
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            1,
            4
          ]
        }
      ],
      "uncertainty": "Samples cannot verify uninterrupted tool execution or the claimed comparative superiority.",
      "media": {
        "sha256": "3017a93536cf7054ccfb49c5c8e3b4f180cd37a3c9251725c03a54fe1135b762",
        "duration": 203.274,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:03.041783+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 601,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "3017a93536cf7054ccfb49c5c8e3b4f180cd37a3c9251725c03a54fe1135b762"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "5iu0nY9lIy8",
      "rank": 3,
      "title": "Use ChatGPT Work to analyze ad performance and refine creative",
      "source_url": "https://www.youtube.com/watch?v=5iu0nY9lIy8",
      "publication_date": "2026-09-04",
      "publication_timestamp": "2026-09-04T05:29:12-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "The speakers connect shopping questions to an advertiser's need to improve campaigns. Reporting develops into interpretation and recommendations, without showing whether those recommendations actually improve performance.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and complete unverified ASR; analytical windows, not cuts. No playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 35,
          "role": "context",
          "summary": "A headphone shopper supplies purchase context.",
          "visual": "Opening card, mobile prompt, headphone advertisement.",
          "evidence_time": 10.5,
          "mechanics": "Prompt and answer views share green margins.",
          "asr_anchor": {
            "word_index": 4,
            "source_time": 4.16
          }
        },
        {
          "start": 35,
          "end": 74,
          "role": "qualification",
          "summary": "Speakers assert separation between ads and answers.",
          "visual": "Headphone ad sits beneath answer controls.",
          "evidence_time": 50.5,
          "mechanics": "Repeated samples preserve the answer/ad spatial separation.",
          "asr_anchor": {
            "word_index": 224,
            "source_time": 36.58
          }
        },
        {
          "start": 74,
          "end": 100,
          "role": "action",
          "summary": "A reporting question produces campaign metrics.",
          "visual": "Dark composer, plugin page, three-row performance table.",
          "evidence_time": 90.5,
          "mechanics": "Consumer interface gives way to advertiser reporting.",
          "asr_anchor": {
            "word_index": 468,
            "source_time": 81.2
          }
        },
        {
          "start": 100,
          "end": 130,
          "role": "discussion",
          "summary": "Follow-ups seek explanations and next actions.",
          "visual": "Blue questions sit above analytical responses.",
          "evidence_time": 110.5,
          "mechanics": "Tabular reporting becomes prose and numbered recommendations.",
          "asr_anchor": {
            "word_index": 742,
            "source_time": 111.02
          }
        },
        {
          "start": 130,
          "end": 154.934,
          "role": "close",
          "summary": "Availability remarks close on proposed improvements.",
          "visual": "Recommendations remain before a white dot card.",
          "evidence_time": 140.5,
          "mechanics": "Speaker inset reappears above the retained response.",
          "asr_anchor": {
            "word_index": 934,
            "source_time": 134.02
          }
        }
      ],
      "editing_lesson": {
        "start": 74,
        "end": 130,
        "observation": "A metrics table precedes a diagnostic question and numbered next steps.",
        "inference": "Interpretation: the question sequence converts reporting into a decision story.",
        "adaptation": "Show your campaign table, ask about a discrepancy, then display a testable recommendation.",
        "caution": "Generated explanations do not demonstrate causation or an improved campaign.",
        "evidence_times": [
          90.5,
          110.5,
          120.5
        ]
      },
      "codes": [
        {
          "code": "question-led",
          "beat_indices": [
            2,
            3
          ]
        },
        {
          "code": "context-before-detail",
          "beat_indices": [
            0,
            2
          ]
        }
      ],
      "uncertainty": "Ad-independence and market-availability statements are attributed claims, not verified by this interface recording.",
      "media": {
        "sha256": "2382e6c92cb147e982735089d936ba179184a865bfabb80196044b28e39d2db4",
        "duration": 154.934,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:15:59.536426+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 513,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "2382e6c92cb147e982735089d936ba179184a865bfabb80196044b28e39d2db4"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "D-QteDHdHes",
      "rank": 4,
      "title": "Use ChatGPT Work to build custom creative tools",
      "source_url": "https://www.youtube.com/watch?v=D-QteDHdHes",
      "publication_date": "2026-09-04",
      "publication_timestamp": "2026-09-04T05:27:16-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A non-coder describes a fictional launch with constrained creative resources. Reference materials lead to a layout tool, asset gallery and website, supporting his account of making production materials himself.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and complete unverified ASR; analytical windows, not cuts. No playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 34,
          "role": "hook",
          "summary": "Finished work introduces a fictional launch problem.",
          "visual": "Colorful headphone website, orange product art, editor.",
          "evidence_time": 15.5,
          "mechanics": "Outputs appear before their construction inputs.",
          "asr_anchor": {
            "word_index": 122,
            "source_time": 16.46
          }
        },
        {
          "start": 34,
          "end": 73,
          "role": "task",
          "summary": "Brand references and wireframe constrain the tool.",
          "visual": "Attachment prompt, brand PDF, hand-drawn interface.",
          "evidence_time": 65.5,
          "mechanics": "Text requirements precede reference-document and sketch views.",
          "asr_anchor": {
            "word_index": 364,
            "source_time": 46.0
          }
        },
        {
          "start": 73,
          "end": 105,
          "role": "proof",
          "summary": "Layout controls lead to multiple creative examples.",
          "visual": "Headphone editor and rows of generated advertisements.",
          "evidence_time": 90.5,
          "mechanics": "Headphone scale/orientation differs before the gallery appears.",
          "asr_anchor": {
            "word_index": 630,
            "source_time": 77.02
          }
        },
        {
          "start": 105,
          "end": 133,
          "role": "task",
          "summary": "A second request reuses assets for a website.",
          "visual": "Website sketch, attachment prompt, headphone image pack.",
          "evidence_time": 120.5,
          "mechanics": "The film returns to inputs before another result.",
          "asr_anchor": {
            "word_index": 918,
            "source_time": 110.46
          }
        },
        {
          "start": 133,
          "end": 166.714,
          "role": "close",
          "summary": "Website examples close the maker's account.",
          "visual": "Colorful homepage, product sections, then publisher mark.",
          "evidence_time": 145.5,
          "mechanics": "Different website sections precede the homepage return.",
          "asr_anchor": {
            "word_index": 1146,
            "source_time": 137.74
          }
        }
      ],
      "editing_lesson": {
        "start": 60,
        "end": 101,
        "observation": "A three-column sketch precedes a corresponding editor and a multi-asset gallery.",
        "inference": "Interpretation: layout continuity connects specification to practical output.",
        "adaptation": "Show your sketch, demonstrate matching controls, then compare assets made with them.",
        "caution": "The gallery is pre-generated; no complete generation cycle is sampled.",
        "evidence_times": [
          65.5,
          75.5,
          90.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            2,
            3,
            4
          ]
        },
        {
          "code": "output-montage",
          "beat_indices": [
            2,
            4
          ]
        }
      ],
      "uncertainty": "Build duration and interactive correctness are asserted, not established by changed stills.",
      "media": {
        "sha256": "75e0faf0b5be70a8cccb6e03042924acc4a742181d88833353fa0aaea8b7034e",
        "duration": 166.714,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:15:59.284936+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 668,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "75e0faf0b5be70a8cccb6e03042924acc4a742181d88833353fa0aaea8b7034e"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "Ez-anO32D_s",
      "rank": 5,
      "title": "Use ChatGPT Images to explore campaign concepts",
      "source_url": "https://www.youtube.com/watch?v=Ez-anO32D_s",
      "publication_date": "2026-09-04",
      "publication_timestamp": "2026-09-04T05:25:30-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "The speakers ask how ordinary billboard locations might support more imaginative campaigns. A scouting image becomes a frontal working view, alternatives and a site composite, ending with the value of exploring more options.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and complete unverified ASR; analytical windows, not cuts. No playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 39,
          "role": "hook",
          "summary": "Campaign examples establish the visual ambition.",
          "visual": "Three building treatments, corridor display, street views.",
          "evidence_time": 20.5,
          "mechanics": "Grouped examples precede a single building's examination.",
          "asr_anchor": {
            "word_index": 110,
            "source_time": 19.0
          }
        },
        {
          "start": 39,
          "end": 68,
          "role": "action",
          "summary": "An angled location photo becomes a working reference.",
          "visual": "Blank billboard photo; frontal facade in chat.",
          "evidence_time": 55.5,
          "mechanics": "Oblique street framing contrasts with rectangular elevation.",
          "asr_anchor": {
            "word_index": 310,
            "source_time": 49.5
          }
        },
        {
          "start": 68,
          "end": 94,
          "role": "proof",
          "summary": "Prompts explore portals and elevated-train concepts.",
          "visual": "Portal variations followed by train-front alternatives.",
          "evidence_time": 85.5,
          "mechanics": "Repeated facade framing holds different concepts comparable.",
          "asr_anchor": {
            "word_index": 480,
            "source_time": 70.06
          }
        },
        {
          "start": 94,
          "end": 120.654,
          "role": "close",
          "summary": "A site composite supports broader exploration.",
          "visual": "Train concept on angled building; publisher mark.",
          "evidence_time": 100.5,
          "mechanics": "Street context replaces the isolated variant grid.",
          "asr_anchor": {
            "word_index": 694,
            "source_time": 95.06
          }
        }
      ],
      "editing_lesson": {
        "start": 70,
        "end": 116,
        "observation": "Portal and train grids precede the selected train treatment in street context.",
        "inference": "Interpretation: contextual return answers whether a concept fits its location.",
        "adaptation": "Compare original concepts in one template, then composite a choice into your location photo.",
        "caution": "The closing image is presented as a composite, not installation evidence.",
        "evidence_times": [
          75.5,
          85.5,
          95.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            2,
            3
          ]
        }
      ],
      "uncertainty": "Samples omit intervening revisions; visual plausibility does not validate building proportions.",
      "media": {
        "sha256": "3d16de6dc5936e8c9dd51245d5c82fac249077dc403753195b903a380c8d76c6",
        "duration": 120.654,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:15:59.036285+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 435,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "3d16de6dc5936e8c9dd51245d5c82fac249077dc403753195b903a380c8d76c6"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "N-MJ1W8Vj9E",
      "rank": 6,
      "title": "Use ChatGPT Work to create campaign emails",
      "source_url": "https://www.youtube.com/watch?v=N-MJ1W8Vj9E",
      "publication_date": "2026-09-04",
      "publication_timestamp": "2026-09-04T05:23:37-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "The speaker describes delays between an email brief and creative handoff. Use-case libraries and packaged skills lead back to finished email layouts, leaving the claimed time savings unmeasured.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and complete unverified ASR; analytical windows, not cuts. No playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 33,
          "role": "hook",
          "summary": "Email examples introduce a shorter handoff claim.",
          "visual": "Four finished email columns and placeholder templates.",
          "evidence_time": 20.5,
          "mechanics": "Populated and placeholder layouts provide a visible contrast.",
          "asr_anchor": {
            "word_index": 6,
            "source_time": 4.019
          }
        },
        {
          "start": 33,
          "end": 70,
          "role": "action",
          "summary": "Spreadsheet content becomes a browsable use-case library.",
          "visual": "Spreadsheet, category tiles, expanded prompt drawer.",
          "evidence_time": 60.5,
          "mechanics": "Dense rows become filtered cards and one detail panel.",
          "asr_anchor": {
            "word_index": 232,
            "source_time": 34.08
          }
        },
        {
          "start": 70,
          "end": 103,
          "role": "proof",
          "summary": "Public examples pair videos with reusable prompts.",
          "visual": "Phone-shaped video panel beside explanatory text.",
          "evidence_time": 90.5,
          "mechanics": "Different video subjects occupy the same page layout.",
          "asr_anchor": {
            "word_index": 608,
            "source_time": 78.4
          }
        },
        {
          "start": 103,
          "end": 145,
          "role": "task",
          "summary": "Packaged skills frame an email-generation request.",
          "visual": "Plugin skill list and populated composer.",
          "evidence_time": 135.5,
          "mechanics": "Capability inventory precedes the concrete request.",
          "asr_anchor": {
            "word_index": 866,
            "source_time": 112.48
          }
        },
        {
          "start": 145,
          "end": 167.514,
          "role": "close",
          "summary": "Finished layouts redirect discussion toward refinement.",
          "visual": "Chat view, email columns, illustrated publisher card.",
          "evidence_time": 150.5,
          "mechanics": "The opening artifact returns after its workflow explanation.",
          "asr_anchor": {
            "word_index": 1166,
            "source_time": 150.46
          }
        }
      ],
      "editing_lesson": {
        "start": 105,
        "end": 161,
        "observation": "Skill inventory precedes an email request and four rendered email columns.",
        "inference": "Interpretation: reusable setup gives the final layouts an intelligible origin.",
        "adaptation": "Show your reusable constraints, submit one brief, and compare the resulting layouts.",
        "caution": "The request specifies five emails, but these sampled views show four columns.",
        "evidence_times": [
          115.5,
          135.5,
          150.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            3,
            4
          ]
        }
      ],
      "uncertainty": "No complete build or email delivery is shown; reported turnaround remains testimonial.",
      "media": {
        "sha256": "bdd067c250d484596c76c5ad10322902f18c3793b698100e5b03f529b680314f",
        "duration": 167.514,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:15:59.288291+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 632,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "bdd067c250d484596c76c5ad10322902f18c3793b698100e5b03f529b680314f"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "HHt8CPJRviA",
      "rank": 7,
      "title": "Use ChatGPT Work to pressure-test marketing campaign briefs",
      "source_url": "https://www.youtube.com/watch?v=HHt8CPJRviA",
      "publication_date": "2026-09-04",
      "publication_timestamp": "2026-09-04T05:21:56-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A deadline makes conventional campaign research difficult. A mock brief passes through simulated buyer reviews and revision, with the closing discussion positioning this as preparation for real research.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and complete unverified ASR; analytical windows, not cuts. No playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 38,
          "role": "context",
          "summary": "A demo brief exposes the research problem.",
          "visual": "Document overview followed by messaging tables.",
          "evidence_time": 25.5,
          "mechanics": "Broad document view precedes dense audience-specific rows.",
          "asr_anchor": {
            "word_index": 46,
            "source_time": 9.22
          }
        },
        {
          "start": 38,
          "end": 75,
          "role": "task",
          "summary": "A skill request specifies synthetic buyer perspectives.",
          "visual": "Long blue prompt lists executive roles.",
          "evidence_time": 45.5,
          "mechanics": "Different crops emphasize one instruction block.",
          "asr_anchor": {
            "word_index": 288,
            "source_time": 40.42
          }
        },
        {
          "start": 75,
          "end": 109,
          "role": "proof",
          "summary": "Simulated reviews produce scores and disagreement.",
          "visual": "Review request, persona table, discussion paragraphs.",
          "evidence_time": 100.5,
          "mechanics": "Structured scorecards precede attributed synthetic dialogue.",
          "asr_anchor": {
            "word_index": 580,
            "source_time": 80.2
          }
        },
        {
          "start": 109,
          "end": 146.594,
          "role": "qualification",
          "summary": "Revision is framed as preparation for fieldwork.",
          "visual": "Revision request and download link; closing mark.",
          "evidence_time": 120.5,
          "mechanics": "Revised-output view persists through the research caveat.",
          "asr_anchor": {
            "word_index": 892,
            "source_time": 120.14
          }
        }
      ],
      "editing_lesson": {
        "start": 75,
        "end": 119,
        "observation": "Persona scorecards precede debate text and a brief-revision request.",
        "inference": "Interpretation: disagreement supplies a reason for the ensuing edit.",
        "adaptation": "Show labeled simulated objections, select a specific change, then reveal the revised passage.",
        "caution": "Synthetic agreement is not customer validation; preserve the field-research qualification.",
        "evidence_times": [
          90.5,
          100.5,
          110.5
        ]
      },
      "codes": [
        {
          "code": "context-before-detail",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            2,
            3
          ]
        }
      ],
      "uncertainty": "Samples do not establish that real customer records informed this explicitly demonstrative exercise.",
      "media": {
        "sha256": "41a99fc205355c19eda00f6a60e0cc549ac1ad32b91f2c10bff671d5b3b19f56",
        "duration": 146.594,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:16:00.917361+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 521,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "41a99fc205355c19eda00f6a60e0cc549ac1ad32b91f2c10bff671d5b3b19f56"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "NSqlZ1L81nQ",
      "rank": 8,
      "title": "Use ChatGPT Work to turn marketing ideas into finished work",
      "source_url": "https://www.youtube.com/watch?v=NSqlZ1L81nQ",
      "publication_date": "2026-09-04",
      "publication_timestamp": "2026-09-04T05:20:19-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Three speakers describe ideas arriving faster from more colleagues. An interface overview explains their proposed execution tools, but the ending shifts responsibility back to marketers' judgment rather than demonstrating business results.",
      "family": "talk",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and complete unverified ASR; analytical windows, not cuts. No playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 42,
          "role": "context",
          "summary": "Speakers describe a shorter idea-to-output gap.",
          "visual": "Three-person table, paired-column diagram, product card.",
          "evidence_time": 20.5,
          "mechanics": "Diagram interrupts and then returns to the conversation.",
          "asr_anchor": {
            "word_index": 8,
            "source_time": 4.16
          }
        },
        {
          "start": 42,
          "end": 80,
          "role": "action",
          "summary": "Mode selection introduces connected workplace tools.",
          "visual": "Two composer states and a plugin directory.",
          "evidence_time": 50.5,
          "mechanics": "Larger composer precedes categorized integration lists.",
          "asr_anchor": {
            "word_index": 310,
            "source_time": 45.12
          }
        },
        {
          "start": 80,
          "end": 105,
          "role": "discussion",
          "summary": "A capability overview broadens the proposition.",
          "visual": "Eight illustrated work categories in two rows.",
          "evidence_time": 90.5,
          "mechanics": "Repeated overview replaces detailed directory browsing.",
          "asr_anchor": {
            "word_index": 568,
            "source_time": 82.18
          }
        },
        {
          "start": 105,
          "end": 154.654,
          "role": "close",
          "summary": "Speakers emphasize judgment and customer relevance.",
          "visual": "Table conversation, plain thesis slide, closing dot.",
          "evidence_time": 130.5,
          "mechanics": "People reappear around the text-only qualification.",
          "asr_anchor": {
            "word_index": 880,
            "source_time": 126.24
          }
        }
      ],
      "editing_lesson": {
        "start": 80,
        "end": 141,
        "observation": "A dense capability grid gives way to people and one plain-text judgment statement.",
        "inference": "Interpretation: reduced visual density distinguishes responsibility from feature breadth.",
        "adaptation": "Follow your feature overview with a single decision criterion and a presenter explaining it.",
        "caution": "The overview demonstrates categories, not successful end-to-end execution.",
        "evidence_times": [
          90.5,
          105.5,
          120.5
        ]
      },
      "codes": [
        {
          "code": "context-before-detail",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            3
          ]
        }
      ],
      "uncertainty": "Directory visibility proves neither authorization nor working integrations; no permission flow is sampled.",
      "media": {
        "sha256": "e8fe2a6da50d03c3755af8a9c98d555ea162ee27dad302dc88e295505f014213",
        "duration": 154.654,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:16:01.314471+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 533,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "e8fe2a6da50d03c3755af8a9c98d555ea162ee27dad302dc88e295505f014213"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "-TTyyY3VWh8",
      "rank": 9,
      "title": "First impressions of GPT-6 Astra from developers",
      "source_url": "https://www.youtube.com/watch?v=-TTyyY3VWh8",
      "publication_date": "2026-09-03",
      "publication_timestamp": "2026-09-03T14:20:49-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "An interviewer asks developers what surprised them. A playable-city example, contrasting web designs and puzzle anecdotes accumulate into enthusiasm, while the puzzle segment explicitly acknowledges outside help.",
      "family": "testimonial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and complete unverified ASR; analytical windows, not cuts. No playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 49,
          "role": "hook",
          "summary": "A question leads into a city simulation.",
          "visual": "Voxel waterfront, street characters, historical timeline controls.",
          "evidence_time": 25.5,
          "mechanics": "Aerial and street-level samples show different city states.",
          "asr_anchor": {
            "word_index": 78,
            "source_time": 10.96
          }
        },
        {
          "start": 49,
          "end": 83,
          "role": "proof",
          "summary": "A designer values unexpected visual alternatives.",
          "visual": "Circular matcha bowl design; draped-green alternative.",
          "evidence_time": 70.5,
          "mechanics": "Contrasting layouts retain the same guest inset.",
          "asr_anchor": {
            "word_index": 318,
            "source_time": 53.1
          }
        },
        {
          "start": 83,
          "end": 112,
          "role": "qualification",
          "summary": "Puzzle success is qualified by creator-provided help.",
          "visual": "Curly-haired guest, cube array, answer page.",
          "evidence_time": 100.5,
          "mechanics": "Conversation precedes enlarged puzzle evidence.",
          "asr_anchor": {
            "word_index": 572,
            "source_time": 88.06
          }
        },
        {
          "start": 112,
          "end": 128,
          "role": "discussion",
          "summary": "Guest attributes persistence to parallel investigation.",
          "visual": "Branch dashboard with guest inset; guest close-up.",
          "evidence_time": 115.5,
          "mechanics": "Dashboard returns to the person interpreting it.",
          "asr_anchor": {
            "word_index": 778,
            "source_time": 114.04
          }
        },
        {
          "start": 128,
          "end": 146.774,
          "role": "close",
          "summary": "Another developer describes clearing previously blocked work.",
          "visual": "Glasses-wearing guest, interviewer, incomplete sampled wordmark.",
          "evidence_time": 140.5,
          "mechanics": "Two-shot narrows to guest before the end card.",
          "asr_anchor": {
            "word_index": 916,
            "source_time": 131.14
          }
        }
      ],
      "editing_lesson": {
        "start": 50,
        "end": 83,
        "observation": "A centered bowl composition contrasts with green drapery before the designer's close-up.",
        "inference": "Interpretation: difference supplies evidence for his account of creative surprise.",
        "adaptation": "Compare two original responses to one brief, then explain the useful difference on camera.",
        "caution": "Two selected outputs reveal neither prompt equivalence nor overall success rate.",
        "evidence_times": [
          55.5,
          70.5,
          75.5
        ]
      },
      "codes": [
        {
          "code": "question-led",
          "beat_indices": [
            0
          ]
        },
        {
          "code": "output-montage",
          "beat_indices": [
            0,
            1,
            2
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            1,
            3
          ]
        }
      ],
      "uncertainty": "City-state samples cannot establish continuous gameplay, historical accuracy or unchanged prompts.",
      "media": {
        "sha256": "9875ab17a5e1e43e2e3d1a84cecc866fa792c71440232f6697641bb00a704a1e",
        "duration": 146.774,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:01.919775+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 499,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "9875ab17a5e1e43e2e3d1a84cecc866fa792c71440232f6697641bb00a704a1e"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "1QNsdr-Qx_I",
      "rank": 10,
      "title": "Introducing GPT-6 Astra: the most intelligent and aligned model in the world.",
      "source_url": "https://www.youtube.com/watch?v=1QNsdr-Qx_I",
      "publication_date": "2026-09-03",
      "publication_timestamp": "2026-09-03T12:40:03-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A simple drawing request expands into increasingly elaborate artifacts while other people delegate everyday work. Recurring tasks receive refinements, and a physical rocket closes the visual thread without establishing its manufacture.",
      "family": "product-film",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and complete unverified ASR; analytical windows, not cuts. No playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 43,
          "role": "task",
          "summary": "A circle request develops into a rocket.",
          "visual": "Projected screens show simple and detailed rockets.",
          "evidence_time": 30.5,
          "mechanics": "User close-ups alternate with wall-scale artifacts.",
          "asr_anchor": {
            "word_index": 30,
            "source_time": 19.04
          }
        },
        {
          "start": 43,
          "end": 80,
          "role": "action",
          "summary": "Parallel requests span presentation, selling and play.",
          "visual": "Different users face presentation, marketplace and Blender screens.",
          "evidence_time": 75.5,
          "mechanics": "Shared room framing links otherwise distinct tasks.",
          "asr_anchor": {
            "word_index": 144,
            "source_time": 47.04
          }
        },
        {
          "start": 80,
          "end": 111,
          "role": "action",
          "summary": "Drafting and scheduling accompany additional listing context.",
          "visual": "Standing user, yellow slides, orange-table detail photograph.",
          "evidence_time": 110.5,
          "mechanics": "Task images and speakers recur in interleaved samples.",
          "asr_anchor": {
            "word_index": 384,
            "source_time": 83.28
          }
        },
        {
          "start": 111,
          "end": 143,
          "role": "proof",
          "summary": "Users refine a draft and presentation styling.",
          "visual": "Highlighted document, space game, slide-background controls.",
          "evidence_time": 135.5,
          "mechanics": "Document selection contrasts with game and color-editing states.",
          "asr_anchor": {
            "word_index": 574,
            "source_time": 112.2
          }
        },
        {
          "start": 143,
          "end": 162.894,
          "role": "close",
          "summary": "Print-file request resolves into a physical rocket image.",
          "visual": "Projected white rocket; tabletop rocket beneath closing cards.",
          "evidence_time": 155.5,
          "mechanics": "Digital and physical rocket forms occupy successive samples.",
          "asr_anchor": {
            "word_index": 718,
            "source_time": 146.0
          }
        }
      ],
      "editing_lesson": {
        "start": 19,
        "end": 162.894,
        "observation": "The rocket recurs as a flat drawing, Blender object and tabletop object.",
        "inference": "Interpretation: one recognizable shape connects escalating tasks across interruptions.",
        "adaptation": "Follow an original motif from sketch through model to a documented physical prototype.",
        "caution": "No printing process is sampled; the physical object could be a prop.",
        "evidence_times": [
          25.5,
          75.5,
          155.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            0,
            1,
            4
          ]
        },
        {
          "code": "output-montage",
          "beat_indices": [
            1,
            2,
            3
          ]
        }
      ],
      "uncertainty": "ASR requests include lawyer review; contract adequacy, bookings and task completion are not independently verified.",
      "media": {
        "sha256": "fe61de3d28d6afcfdd09086e88d8fe349458593f74ac9a240513a6d37b3c93e5",
        "duration": 162.894,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:01.463106+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 374,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "fe61de3d28d6afcfdd09086e88d8fe349458593f74ac9a240513a6d37b3c93e5"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "yfnh8SuMbNw",
      "rank": 11,
      "title": "Use ChatGPT Work on Your Phone",
      "source_url": "https://www.youtube.com/watch?v=yfnh8SuMbNw",
      "publication_date": "2026-08-28",
      "publication_timestamp": "2026-08-28T13:51:46-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Catching up competes with time away from a desk. Mobile summaries expand into finances and remote access, with computer availability qualifying the convenience.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; analytical windows, not cuts or playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 17,
          "role": "context",
          "summary": "Presenter introduces catching up before returning.",
          "visual": "Phone held above laptop; presenter close-up.",
          "evidence_time": 5.5,
          "mechanics": "Wide view establishes devices before tighter framing.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 3.94
          }
        },
        {
          "start": 17,
          "end": 43,
          "role": "proof",
          "summary": "Dictated request becomes a work recap.",
          "visual": "Waveform, then tall phone response with links.",
          "evidence_time": 35.5,
          "mechanics": "Phone dominates; presenter remains a small inset.",
          "asr_anchor": {
            "word_index": 128,
            "source_time": 17.08
          }
        },
        {
          "start": 43,
          "end": 75,
          "role": "action",
          "summary": "Finance review adds a recurring-check request.",
          "visual": "Presenter, phone balance summary, enlarged spending detail.",
          "evidence_time": 65.5,
          "mechanics": "Second waveform precedes a different response topic.",
          "asr_anchor": {
            "word_index": 344,
            "source_time": 43.18
          }
        },
        {
          "start": 75,
          "end": 114.781333,
          "role": "qualification",
          "summary": "Remote access requires an available computer; presenter concludes.",
          "visual": "Desktop chat, paired phone view, connection settings, presenter.",
          "evidence_time": 95.5,
          "mechanics": "Dual-device composition precedes settings and presenter return.",
          "asr_anchor": {
            "word_index": 622,
            "source_time": 75.02
          }
        }
      ],
      "editing_lesson": {
        "start": 80,
        "end": 103,
        "observation": "Desktop chat becomes a phone-and-desktop comparison, then connection settings.",
        "inference": "Interpretation: the dependency follows the convenience claim before the recap.",
        "adaptation": "Show an original desktop task, its phone view, then the required availability setting.",
        "caution": "These samples do not show the narrated pairing-code scan.",
        "evidence_times": [
          90.5,
          95.5,
          100.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            2,
            3
          ]
        }
      ],
      "uncertainty": "Read-only finance access is narrated, not independently tested here.",
      "media": {
        "sha256": "ae53ae5197f5869b0a1c38e6182edb6f87e80328b8f4eaaaaf16e960f96b9793",
        "duration": 114.781333,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:01.921977+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 459,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "ae53ae5197f5869b0a1c38e6182edb6f87e80328b8f4eaaaaf16e960f96b9793"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "syML5KT-HzI",
      "rank": 12,
      "title": "Create Slides, Docs, and Templates",
      "source_url": "https://www.youtube.com/watch?v=syML5KT-HzI",
      "publication_date": "2026-08-28",
      "publication_timestamp": "2026-08-28T13:51:46-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A sprawling meeting transcript must serve two audiences. Document and slide outputs establish the transformation before saved templates extend it to another week.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; analytical windows, not cuts or playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 44,
          "role": "context",
          "summary": "One meeting needs two useful formats.",
          "visual": "Dense transcript panel beside seated presenter.",
          "evidence_time": 10.5,
          "mechanics": "Text overlay makes source volume visible beside explanation.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 3.58
          }
        },
        {
          "start": 44,
          "end": 100,
          "role": "action",
          "summary": "Connect Drive and specify the recap structure.",
          "visual": "Connection dialog, then multiline prompt with headings.",
          "evidence_time": 55.5,
          "mechanics": "Account setup precedes progressively longer prompt samples.",
          "asr_anchor": {
            "word_index": 248,
            "source_time": 44.46
          }
        },
        {
          "start": 100,
          "end": 150,
          "role": "proof",
          "summary": "Document detail becomes leadership slides.",
          "visual": "Action-item table, then three-card slide.",
          "evidence_time": 145.5,
          "mechanics": "Document and deck occupy comparable large viewing areas.",
          "asr_anchor": {
            "word_index": 484,
            "source_time": 100.1
          }
        },
        {
          "start": 150,
          "end": 188,
          "role": "action",
          "summary": "Save formats and demonstrate another week's outputs.",
          "visual": "Template interface, later document and slide previews.",
          "evidence_time": 180.5,
          "mechanics": "Reuse request precedes another paired output sequence.",
          "asr_anchor": {
            "word_index": 778,
            "source_time": 150.1
          }
        },
        {
          "start": 188,
          "end": 197.074,
          "role": "close",
          "summary": "Presenter recaps reduced rebuilding.",
          "visual": "Seated presenter, then white logo card.",
          "evidence_time": 190.5,
          "mechanics": "Interface disappears before the brand-only sample.",
          "asr_anchor": {
            "word_index": 974,
            "source_time": 188.0
          }
        }
      ],
      "editing_lesson": {
        "start": 100,
        "end": 150,
        "observation": "An action-item table precedes a summary document and three-card slide.",
        "inference": "Interpretation: contrasting containers explains audience-specific compression.",
        "adaptation": "Show your meeting's detailed actions, summary, then decision slide using matching facts.",
        "caution": "The samples cannot establish every slide's fidelity to the transcript.",
        "evidence_times": [
          110.5,
          140.5,
          145.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            2,
            3
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            4
          ]
        }
      ],
      "uncertainty": "Template reuse is demonstrated selectively, not across complete generation runs.",
      "media": {
        "sha256": "f7428f1b81875b066182aaaa56c50ca7d1d803c4dbe178ae3223400e505457de",
        "duration": 197.074,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:03.349891+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 498,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "f7428f1b81875b066182aaaa56c50ca7d1d803c4dbe178ae3223400e505457de"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "DLcQSp_CIIc",
      "rank": 13,
      "title": "Talk to ChatGPT Work",
      "source_url": "https://www.youtube.com/watch?v=DLcQSp_CIIc",
      "publication_date": "2026-08-28",
      "publication_timestamp": "2026-08-28T13:51:46-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Can a spoken brainstorm produce something usable? Conversation over a desktop workspace leads to a pitch deck, followed by a phone-access qualification.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; analytical windows, not cuts or playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 18,
          "role": "hook",
          "summary": "Presenter requests help with tasks and music.",
          "visual": "Overhead room, desktop Notes window, presenter.",
          "evidence_time": 5.5,
          "mechanics": "Voice orb appears over an existing workspace.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 3.36
          }
        },
        {
          "start": 18,
          "end": 63,
          "role": "discussion",
          "summary": "Brainstorm explores participatory event concepts.",
          "visual": "Dark Notes window over browser; purple orb.",
          "evidence_time": 35.5,
          "mechanics": "Similar workspace persists across several conversational samples.",
          "asr_anchor": {
            "word_index": 124,
            "source_time": 18.3
          }
        },
        {
          "start": 63,
          "end": 94,
          "role": "proof",
          "summary": "Deck request leads to illustrated concepts.",
          "visual": "Presenter, then slides depicting installations and devices.",
          "evidence_time": 85.5,
          "mechanics": "Orb remains visible across workspace and slide samples.",
          "asr_anchor": {
            "word_index": 402,
            "source_time": 63.0
          }
        },
        {
          "start": 94,
          "end": 110.902458,
          "role": "qualification",
          "summary": "Phone access is explained before the recap.",
          "visual": "App menu, presenter holding phone, logo.",
          "evidence_time": 100.5,
          "mechanics": "Phone demonstration returns attention to the presenter.",
          "asr_anchor": {
            "word_index": 626,
            "source_time": 94.14
          }
        }
      ],
      "editing_lesson": {
        "start": 63,
        "end": 94,
        "observation": "The voice orb appears over Notes and later over illustrated slides.",
        "inference": "Interpretation: a recurring interface element connects conversation to artifact.",
        "adaptation": "Keep your genuine voice indicator visible while revealing the resulting original deck.",
        "caution": "The missing generation interval prevents real-time speed claims.",
        "evidence_times": [
          70.5,
          80.5,
          90.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "code": "proof-before-close",
          "beat_indices": [
            2,
            3
          ]
        }
      ],
      "uncertainty": "ASR product-name spellings are unreliable; spoken responses were not listened to.",
      "media": {
        "sha256": "370a752913021cb1d279b29e7c52e270203229bc3f238e11f6f27f8352820b66",
        "duration": 110.902458,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:03.496632+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 360,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "370a752913021cb1d279b29e7c52e270203229bc3f238e11f6f27f8352820b66"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "981SivztzOc",
      "rank": 14,
      "title": "Use Your Computer and Browser",
      "source_url": "https://www.youtube.com/watch?v=981SivztzOc",
      "publication_date": "2026-08-28",
      "publication_timestamp": "2026-08-28T13:51:46-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Some tasks sit outside connected integrations. A music example introduces computer interaction before a leave-request example makes human review the endpoint.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; analytical windows, not cuts or playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 42,
          "role": "context",
          "summary": "Presenter distinguishes access routes and permissions.",
          "visual": "Presenter alternates with plugin catalog.",
          "evidence_time": 15.5,
          "mechanics": "Catalog insert sits between presenter-led explanations.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 4.86
          }
        },
        {
          "start": 42,
          "end": 78,
          "role": "action",
          "summary": "Music request precedes an app-context example.",
          "visual": "Spotify preview, webpage, attached screenshot in composer.",
          "evidence_time": 70.5,
          "mechanics": "Desktop context becomes a small input attachment.",
          "asr_anchor": {
            "word_index": 280,
            "source_time": 42.5
          }
        },
        {
          "start": 78,
          "end": 112,
          "role": "task",
          "summary": "Leave planning includes a stop-before-submission instruction.",
          "visual": "Dictation waveform, chat response with browser preview.",
          "evidence_time": 100.5,
          "mechanics": "Waveform samples precede the response-and-preview arrangement.",
          "asr_anchor": {
            "word_index": 542,
            "source_time": 78.02
          }
        },
        {
          "start": 112,
          "end": 130,
          "role": "proof",
          "summary": "Prepared drafts remain subject to review.",
          "visual": "Chat summary and filled absence-request modal.",
          "evidence_time": 125.5,
          "mechanics": "Form exposes its submission button after the summary.",
          "asr_anchor": {
            "word_index": 778,
            "source_time": 112.44
          }
        },
        {
          "start": 130,
          "end": 148.914,
          "role": "close",
          "summary": "Presenter restates tool choice and retained control.",
          "visual": "Medium and tight presenter views; white end card.",
          "evidence_time": 135.5,
          "mechanics": "Presenter replaces the unfinished transaction interface.",
          "asr_anchor": {
            "word_index": 910,
            "source_time": 130.32
          }
        }
      ],
      "editing_lesson": {
        "start": 112,
        "end": 130,
        "observation": "Summary screens precede a filled form with submission still available.",
        "inference": "Interpretation: the unresolved button makes retained authority visible.",
        "adaptation": "Show your prepared request, inspect its fields, and end before committing it.",
        "caution": "A visible button alone cannot prove no background submission occurred.",
        "evidence_times": [
          115.5,
          120.5,
          125.5
        ]
      },
      "codes": [
        {
          "code": "permission-step",
          "beat_indices": [
            2,
            3
          ]
        },
        {
          "code": "proof-before-close",
          "beat_indices": [
            3,
            4
          ]
        }
      ],
      "uncertainty": "Platform restrictions and background behavior remain the presenter's claims.",
      "media": {
        "sha256": "f0897ef1238ec2b3b76273ce6906af0758dc700b59d8f6b8d38f3ec71b68d486",
        "duration": 148.914,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:03.246906+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 513,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "f0897ef1238ec2b3b76273ce6906af0758dc700b59d8f6b8d38f3ec71b68d486"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "urXc4xxixVU",
      "rank": 15,
      "title": "Scheduled Tasks",
      "source_url": "https://www.youtube.com/watch?v=urXc4xxixVU",
      "publication_date": "2026-08-28",
      "publication_timestamp": "2026-08-28T13:51:45-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Repeated meeting preparation motivates automation. The presenter first checks a single brief, then schedules that task and refines its format.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; analytical windows, not cuts or playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 48,
          "role": "context",
          "summary": "Repeated preparation needs connected calendar and email.",
          "visual": "Presenter, web workspace, plugin catalog.",
          "evidence_time": 45.5,
          "mechanics": "Catalog detail follows broad setup views.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 4.16
          }
        },
        {
          "start": 48,
          "end": 69,
          "role": "proof",
          "summary": "Inspect one brief before scheduling it.",
          "visual": "Meeting prompt and side-by-side brief document.",
          "evidence_time": 65.5,
          "mechanics": "Prompt-only sample precedes a readable output pane.",
          "asr_anchor": {
            "word_index": 230,
            "source_time": 48.58
          }
        },
        {
          "start": 69,
          "end": 104,
          "role": "action",
          "summary": "Schedule the routine and apply saved instructions.",
          "visual": "Scheduled-task list, timing card, expanded brief.",
          "evidence_time": 80.5,
          "mechanics": "Schedule views precede the revised document view.",
          "asr_anchor": {
            "word_index": 326,
            "source_time": 69.6
          }
        },
        {
          "start": 104,
          "end": 129.674,
          "role": "close",
          "summary": "Other recurring requests broaden the example.",
          "visual": "Different prompt cards beside the presenter.",
          "evidence_time": 110.5,
          "mechanics": "Small overlays replace the full-screen workflow.",
          "asr_anchor": {
            "word_index": 490,
            "source_time": 104.08
          }
        }
      ],
      "editing_lesson": {
        "start": 48,
        "end": 84,
        "observation": "Request, completed brief, and schedule listing appear in that order.",
        "inference": "Interpretation: useful output earns repetition rather than assuming it.",
        "adaptation": "Run your task once, inspect the result, then show its recurring configuration.",
        "caution": "The listing does not demonstrate a later unattended run.",
        "evidence_times": [
          55.5,
          65.5,
          80.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1
          ]
        },
        {
          "code": "context-before-detail",
          "beat_indices": [
            0,
            1
          ]
        }
      ],
      "uncertainty": "Future notifications and recurring execution are not observed.",
      "media": {
        "sha256": "401b9d9b874acf71d53dc8378b934109933de3dc7aeb908f87cfac30e40c502e",
        "duration": 129.674,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:03.801648+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 292,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "401b9d9b874acf71d53dc8378b934109933de3dc7aeb908f87cfac30e40c502e"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "5QPBVrAAdBk",
      "rank": 16,
      "title": "Plugins & Skills",
      "source_url": "https://www.youtube.com/watch?v=5QPBVrAAdBk",
      "publication_date": "2026-08-28",
      "publication_timestamp": "2026-08-28T13:51:45-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A client kickoff requires scattered information and a repeatable brief. Calendar and email work produces the first brief before the presenter requests reuse as a skill.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills, one enlarged frame, unverified ASR; windows are not cuts or playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 62,
          "role": "context",
          "summary": "Client details motivate connecting calendar access.",
          "visual": "Busy calendar, client email, plugin detail page.",
          "evidence_time": 15.5,
          "mechanics": "Separate source screens precede integration setup.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 4.78
          }
        },
        {
          "start": 62,
          "end": 118,
          "role": "action",
          "summary": "Review invitation, authorize booking, then prepare brief.",
          "visual": "Availability response, invitation fields, structured brief.",
          "evidence_time": 115.5,
          "mechanics": "Request and response views progressively expose meeting detail.",
          "asr_anchor": {
            "word_index": 334,
            "source_time": 62.68
          }
        },
        {
          "start": 118,
          "end": 150,
          "role": "action",
          "summary": "Save the method and request another run.",
          "visual": "Presenter, skill confirmation, new execution response.",
          "evidence_time": 145.5,
          "mechanics": "Saved-workflow confirmation precedes a separate reuse request.",
          "asr_anchor": {
            "word_index": 590,
            "source_time": 118.08
          }
        },
        {
          "start": 150,
          "end": 165.174,
          "role": "close",
          "summary": "Presenter recaps preparation and reuse.",
          "visual": "Medium presenter, close-up, white logo card.",
          "evidence_time": 160.5,
          "mechanics": "Human recap follows the workflow screens.",
          "asr_anchor": {
            "word_index": 782,
            "source_time": 150.14
          }
        }
      ],
      "editing_lesson": {
        "start": 105,
        "end": 150,
        "observation": "Brief detail precedes skill confirmation and a subsequent run request.",
        "inference": "Interpretation: an existing result explains why saving instructions matters.",
        "adaptation": "Show your accepted brief, save its method, then request a second example.",
        "caution": "The inspected rerun frame shows work starting, not a completed second brief.",
        "evidence_times": [
          115.5,
          140.5,
          145.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            3
          ]
        }
      ],
      "uncertainty": "Narrated reuse success exceeds the second result visible in sampled frames.",
      "media": {
        "sha256": "f8c2c86c9945ecc11f0ca019ecb907bdba3da2131289e5e95bf9fcefd9169e67",
        "duration": 165.174,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:05.005325+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 423,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "f8c2c86c9945ecc11f0ca019ecb907bdba3da2131289e5e95bf9fcefd9169e67"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "0qK3KqvtRrg",
      "rank": 17,
      "title": "Build a Shareable Site",
      "source_url": "https://www.youtube.com/watch?v=0qK3KqvtRrg",
      "publication_date": "2026-08-28",
      "publication_timestamp": "2026-08-28T13:51:45-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Trip details need a shared working destination. Requests produce an expense-bearing site, sharing is discussed, and a launch dashboard extends the premise beyond travel.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; analytical windows, not cuts or playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 32,
          "role": "context",
          "summary": "Existing trip planning motivates a group site.",
          "visual": "Presenter and travel webpage with landscape imagery.",
          "evidence_time": 15.5,
          "mechanics": "Finished-looking webpage appears before the detailed request.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 4.9
          }
        },
        {
          "start": 32,
          "end": 78,
          "role": "action",
          "summary": "Trip information becomes a site with expenses.",
          "visual": "Reservation text, dictation, build preview, expense table.",
          "evidence_time": 75.5,
          "mechanics": "Text-heavy source gives way to chat-and-site panes.",
          "asr_anchor": {
            "word_index": 226,
            "source_time": 32.22
          }
        },
        {
          "start": 78,
          "end": 99,
          "role": "qualification",
          "summary": "Presenter explains private and shared access.",
          "visual": "Expense balances beside a sharing request.",
          "evidence_time": 90.5,
          "mechanics": "Existing site remains beside the new access instruction.",
          "asr_anchor": {
            "word_index": 536,
            "source_time": 78.12
          }
        },
        {
          "start": 99,
          "end": 140.515375,
          "role": "close",
          "summary": "Dashboard updates extend the workflow; sites index closes.",
          "visual": "Dashboard-linked dictation, Sites list, presenter, logo.",
          "evidence_time": 130.5,
          "mechanics": "Second task ends on a multi-site inventory.",
          "asr_anchor": {
            "word_index": 688,
            "source_time": 99.36
          }
        }
      ],
      "editing_lesson": {
        "start": 32,
        "end": 78,
        "observation": "Reservation text precedes a build-status pane and populated expense interface.",
        "inference": "Interpretation: distinct states make the source-to-tool transformation inspectable.",
        "adaptation": "Show original trip inputs, a genuine build state, then a working expense view.",
        "caution": "A populated table does not establish persistence or correct cost splitting.",
        "evidence_times": [
          35.5,
          55.5,
          75.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            2,
            3
          ]
        }
      ],
      "uncertainty": "Sharing enforcement and future dashboard refreshes remain untested claims.",
      "media": {
        "sha256": "177b48e82d7e6815e04f0d69b6a488080210be0d6595e9b0ec96b33ad9bdc6a5",
        "duration": 140.515375,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:06.678126+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 472,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "177b48e82d7e6815e04f0d69b6a488080210be0d6595e9b0ec96b33ad9bdc6a5"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "Rk4VyQxDq5s",
      "rank": 18,
      "title": "Getting Started with ChatGPT Work",
      "source_url": "https://www.youtube.com/watch?v=Rk4VyQxDq5s",
      "publication_date": "2026-08-28",
      "publication_timestamp": "2026-08-28T13:51:11-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Dinner arrangements are split between email and team messages. Installation and authorization enable a combined draft, with a human edit and send decision resolving the task.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; analytical windows, not cuts or playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 44,
          "role": "task",
          "summary": "Scattered dinner details establish the reason for setup.",
          "visual": "Presenter with mode selector and floating app icons.",
          "evidence_time": 35.5,
          "mechanics": "Gmail and Slack icons externalize the split sources.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 4.16
          }
        },
        {
          "start": 44,
          "end": 123,
          "role": "action",
          "summary": "Install, select mode, and authorize source access.",
          "visual": "Installer, mode switch, Google consent, Slack page.",
          "evidence_time": 110.5,
          "mechanics": "Consent occupies its own screen between setup states.",
          "asr_anchor": {
            "word_index": 238,
            "source_time": 44.14
          }
        },
        {
          "start": 123,
          "end": 171,
          "role": "proof",
          "summary": "Request joins restaurant email and team constraints.",
          "visual": "Prompt, email detail, adjacent Slack conversation.",
          "evidence_time": 165.5,
          "mechanics": "Source windows follow the gathering response.",
          "asr_anchor": {
            "word_index": 636,
            "source_time": 123.68
          }
        },
        {
          "start": 171,
          "end": 189,
          "role": "action",
          "summary": "Edit the reply before authorizing send.",
          "visual": "Email draft and enlarged blue send control.",
          "evidence_time": 185.5,
          "mechanics": "Tighter button framing follows the full draft.",
          "asr_anchor": {
            "word_index": 912,
            "source_time": 171.36
          }
        },
        {
          "start": 189,
          "end": 233.154,
          "role": "close",
          "summary": "Attention alerts broaden the completed example.",
          "visual": "Small desktop pet, presenter, white end card.",
          "evidence_time": 200.5,
          "mechanics": "Pet insert interrupts the concluding presenter sequence.",
          "asr_anchor": {
            "word_index": 1020,
            "source_time": 189.52
          }
        }
      ],
      "editing_lesson": {
        "start": 159,
        "end": 189,
        "observation": "Email and Slack source views precede the draft and enlarged send button.",
        "inference": "Interpretation: provenance comes before the consequential action.",
        "adaptation": "Show your two sources, reconcile their facts in a draft, then expose approval.",
        "caution": "Cursor placement on send is not a delivery receipt.",
        "evidence_times": [
          165.5,
          180.5,
          185.5
        ]
      },
      "codes": [
        {
          "code": "context-before-detail",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "code": "permission-step",
          "beat_indices": [
            1,
            3
          ]
        },
        {
          "code": "input-to-artifact",
          "beat_indices": [
            2,
            3
          ]
        }
      ],
      "uncertainty": "The sampled send control does not independently establish successful delivery.",
      "media": {
        "sha256": "2afb842224820ace57f43bf9e771c3abb632993598e108da208a99b1c27b5c46",
        "duration": 233.154,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:06.782415+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 622,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "2afb842224820ace57f43bf9e771c3abb632993598e108da208a99b1c27b5c46"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "Lj03DNImB3s",
      "rank": 19,
      "title": "Delivering more meals to more moms with ChatGPT",
      "source_url": "https://www.youtube.com/watch?v=Lj03DNImB3s",
      "publication_date": "2026-08-27",
      "publication_timestamp": "2026-08-27T14:30:15-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "shorts",
        "uploads"
      ],
      "narrative": "A meal provider's workload makes label preparation a concrete bottleneck. Kitchen coaching leads to visible label designs and reported time savings, then returns to delivery.",
      "family": "testimonial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second portrait stills, one enlarged frame, unverified ASR; no playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 23,
          "role": "context",
          "summary": "Cooking and administration compete for one person's time.",
          "visual": "Kitchen pair, delivery bag, cook with task-word overlays.",
          "evidence_time": 20.5,
          "mechanics": "Accumulated words surround the worker rather than software.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.0
          }
        },
        {
          "start": 23,
          "end": 43,
          "role": "action",
          "summary": "Coach connects design tools and source information.",
          "visual": "Kitchen laptop below plugin-interface inserts.",
          "evidence_time": 35.5,
          "mechanics": "Portrait split composition retains both people beneath software.",
          "asr_anchor": {
            "word_index": 170,
            "source_time": 23.0
          }
        },
        {
          "start": 43,
          "end": 60,
          "role": "proof",
          "summary": "Labels appear; participants report faster completion.",
          "visual": "Chat, stacked food-label designs, cook's reaction.",
          "evidence_time": 50.5,
          "mechanics": "Artifact occupies the frame before returning to kitchen.",
          "asr_anchor": {
            "word_index": 318,
            "source_time": 43.08
          }
        },
        {
          "start": 60,
          "end": 73.754,
          "role": "close",
          "summary": "Invitation to participate precedes a delivery glimpse.",
          "visual": "Coach at counter, exterior steps, white branding.",
          "evidence_time": 65.5,
          "mechanics": "Domestic work context expands to the customer's doorway.",
          "asr_anchor": {
            "word_index": 424,
            "source_time": 60.1
          }
        }
      ],
      "editing_lesson": {
        "start": 43,
        "end": 60,
        "observation": "Chat gives way to label designs and then the cook beside ingredients.",
        "inference": "Interpretation: returning to physical work supplies a human payoff.",
        "adaptation": "Show an original admin result, then its owner continuing the interrupted task.",
        "caution": "Reported time savings lack an independently observed timer.",
        "evidence_times": [
          45.5,
          50.5,
          55.5
        ]
      },
      "codes": [
        {
          "code": "human-stakes",
          "beat_indices": [
            0,
            2,
            3
          ]
        },
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            2
          ]
        }
      ],
      "uncertainty": "Visible food labels do not verify the narrated spreadsheet-to-name update.",
      "media": {
        "sha256": "27455259ab99c396bae4c39c3fe9c05d991020af0b4e12c783a9f0e859a092c7",
        "duration": 73.754,
        "width": 240,
        "height": 426,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:05.478720+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 250,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "27455259ab99c396bae4c39c3fe9c05d991020af0b4e12c783a9f0e859a092c7"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "t64oZKCdG8Q",
      "rank": 20,
      "title": "How to Build a Personalized Meal Planner with ChatGPT Work",
      "source_url": "https://www.youtube.com/watch?v=t64oZKCdG8Q",
      "publication_date": "2026-08-27",
      "publication_timestamp": "2026-08-27T09:16:32-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A parent's meal-planning burden includes a child's preferences and allergies. Notes become a shared planner and shopping cart, while feedback and clinician-message drafts retain review points.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills, one enlarged frame, unverified ASR; windows are not cuts or playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 24,
          "role": "context",
          "summary": "Presenter frames family planning constraints.",
          "visual": "Laptop opening; medium and tight presenter views.",
          "evidence_time": 15.5,
          "mechanics": "Personal explanation precedes any planning interface.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 2.7
          }
        },
        {
          "start": 24,
          "end": 78,
          "role": "action",
          "summary": "Food notes become plan, site, and grocery cart.",
          "visual": "Screenshot attachment, meal table, colorful site, checkout link.",
          "evidence_time": 75.5,
          "mechanics": "Text plan becomes a side-by-side designed site.",
          "asr_anchor": {
            "word_index": 128,
            "source_time": 24.1
          }
        },
        {
          "start": 78,
          "end": 113,
          "role": "qualification",
          "summary": "Feedback informs reminders and a clinician-message draft.",
          "visual": "Phone rating form, scheduled task, unsent email draft.",
          "evidence_time": 110.5,
          "mechanics": "Phone feedback precedes schedule and correspondence samples.",
          "asr_anchor": {
            "word_index": 396,
            "source_time": 78.06
          }
        },
        {
          "start": 113,
          "end": 127.514,
          "role": "close",
          "summary": "Presenter concludes and takes a personal call.",
          "visual": "Presenter at laptop, then phone at ear, end card.",
          "evidence_time": 120.5,
          "mechanics": "Working posture gives way to a phone-call tableau.",
          "asr_anchor": {
            "word_index": 564,
            "source_time": 113.16
          }
        }
      ],
      "editing_lesson": {
        "start": 78,
        "end": 113,
        "observation": "Phone feedback, recurring-task confirmation, and clinician-message draft form three distinct states.",
        "inference": "Interpretation: the story extends planning into a reviewable feedback loop.",
        "adaptation": "Use a nonmedical example: record feedback, schedule review, then draft a human-approved follow-up.",
        "caution": "A clinician-addressed draft establishes neither medical safety nor diagnostic validity.",
        "evidence_times": [
          90.5,
          100.5,
          110.5
        ]
      },
      "codes": [
        {
          "code": "human-stakes",
          "beat_indices": [
            0,
            3
          ]
        },
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            2
          ]
        }
      ],
      "uncertainty": "Allergy-related suitability is the film's premise, not independently validated advice.",
      "media": {
        "sha256": "9accd72cd0119d7244864cf55b446779dc7ab2916aa8525c741e04f9fcf4497d",
        "duration": 127.514,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:05.426377+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 305,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "9accd72cd0119d7244864cf55b446779dc7ab2916aa8525c741e04f9fcf4497d"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "o38xYi2mtgc",
      "rank": 21,
      "title": "What Codex Unlocks for loveholidays",
      "source_url": "https://www.youtube.com/watch?v=o38xYi2mtgc",
      "publication_date": "2026-08-26",
      "publication_timestamp": "2026-08-26T02:06:31-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Two employees frame AI delegation as changing who can build. Question cards organize company context, self-service examples and claimed gains, resolving in broader participation rather than demonstrated software.",
      "family": "testimonial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; windows are analytical, not cuts; no playback."
      },
      "beats": [
        {
          "start": 0,
          "end": 35,
          "role": "hook",
          "summary": "Delegation introduces the company's business context.",
          "visual": "Navy-shirted respondent on blue sofa; cushion adjustment.",
          "evidence_time": 15.5,
          "mechanics": "Informal setup interrupts seated interview samples.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.08
          }
        },
        {
          "start": 35,
          "end": 95,
          "role": "context",
          "summary": "Respondents describe shared organizational expertise.",
          "visual": "Question cards, second respondent, camera monitor.",
          "evidence_time": 60.5,
          "mechanics": "Cards separate topics; framing ranges from monitor to close-up.",
          "asr_anchor": {
            "word_index": 236,
            "source_time": 36.68
          }
        },
        {
          "start": 95,
          "end": 137,
          "role": "proof",
          "summary": "A respondent describes non-engineers testing search ideas.",
          "visual": "White-shirted respondent beside boom microphone.",
          "evidence_time": 120.5,
          "mechanics": "Wide and tight interview views substitute for software evidence.",
          "asr_anchor": {
            "word_index": 536,
            "source_time": 97.74
          }
        },
        {
          "start": 137,
          "end": 175,
          "role": "proof",
          "summary": "Both respondents attribute operational gains to adoption.",
          "visual": "Each respondent occupies the same sofa setting.",
          "evidence_time": 155.5,
          "mechanics": "Different speakers share matching interview geography.",
          "asr_anchor": {
            "word_index": 722,
            "source_time": 137.6
          }
        },
        {
          "start": 175,
          "end": 212.274,
          "role": "close",
          "summary": "Business impact culminates in broader participation claims.",
          "visual": "Studio lights, final question, shared clapperboard, logo.",
          "evidence_time": 205.5,
          "mechanics": "Separate interviews end with both respondents together.",
          "asr_anchor": {
            "word_index": 934,
            "source_time": 177.02
          }
        }
      ],
      "editing_lesson": {
        "start": 35,
        "end": 50,
        "observation": "A question card precedes wide and tight views of one respondent.",
        "inference": "Interpretation: the question unifies changing interview scales.",
        "adaptation": "Film an original question card, room-wide answer and matching close-up.",
        "caution": "Sample order does not establish continuous speech across angles.",
        "evidence_times": [
          35.5,
          40.5,
          45.5
        ]
      },
      "codes": [
        {
          "code": "question-led",
          "beat_indices": [
            1,
            4
          ]
        }
      ],
      "uncertainty": "Search functionality and numerical gains remain testimony, not inspected product evidence. ASR technical wording is unverified.",
      "media": {
        "sha256": "716c37008e5aef89cb19b21f5d4bb1bb92eafd6f08a60a5d0ce90db7628d02c9",
        "duration": 212.274,
        "width": 854,
        "height": 480,
        "fps": "25/1",
        "retrieved_at": "2026-09-17T18:16:06.341048+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 556,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "716c37008e5aef89cb19b21f5d4bb1bb92eafd6f08a60a5d0ce90db7628d02c9"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "29SyCndnMZs",
      "rank": 22,
      "title": "How to Manage Your Workspace With ChatGPT Work",
      "source_url": "https://www.youtube.com/watch?v=29SyCndnMZs",
      "publication_date": "2026-08-25",
      "publication_timestamp": "2026-08-25T13:17:22-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Fragmented administration is presented as a context-switching problem. Installation, charts, sharing and request approval develop a conversational alternative before the presenter returns to summarize it.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; analytical windows, no listening or moving review."
      },
      "beats": [
        {
          "start": 0,
          "end": 30,
          "role": "context",
          "summary": "Presenter describes fragmented administrative work and existing permissions.",
          "visual": "Woman in green at laptop.",
          "evidence_time": 20.5,
          "mechanics": "Stable table framing precedes screen evidence.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.74
          }
        },
        {
          "start": 30,
          "end": 60,
          "role": "task",
          "summary": "Install plugin and request adoption analysis.",
          "visual": "Plugin directory, composer and dictation waveform.",
          "evidence_time": 45.5,
          "mechanics": "Full desktop gives way to enlarged input.",
          "asr_anchor": {
            "word_index": 152,
            "source_time": 30.16
          }
        },
        {
          "start": 60,
          "end": 98,
          "role": "proof",
          "summary": "Analysis develops into a shared rollout deck.",
          "visual": "Charts, written findings, slide thumbnails, Slack preview.",
          "evidence_time": 90.5,
          "mechanics": "Analytical outputs precede a sharing artifact.",
          "asr_anchor": {
            "word_index": 312,
            "source_time": 60.12
          }
        },
        {
          "start": 98,
          "end": 128,
          "role": "action",
          "summary": "Inspect a limit request before approving it.",
          "visual": "Request table, usage chart, approval confirmation.",
          "evidence_time": 125.5,
          "mechanics": "List narrows to one request and decision.",
          "asr_anchor": {
            "word_index": 546,
            "source_time": 99.36
          }
        },
        {
          "start": 128,
          "end": 143.674,
          "role": "close",
          "summary": "Presenter recaps the consolidated workflow.",
          "visual": "Table presenter returns before white wordmark.",
          "evidence_time": 130.5,
          "mechanics": "Human framing replaces the interface.",
          "asr_anchor": {
            "word_index": 708,
            "source_time": 128.88
          }
        }
      ],
      "editing_lesson": {
        "start": 98,
        "end": 128,
        "observation": "Request rows precede a usage chart and approval confirmation.",
        "inference": "Interpretation: inspection makes the decision appear accountable.",
        "adaptation": "Capture your request list, relevant usage, explicit approval and confirmation.",
        "caution": "The displayed confirmation does not audit underlying permission enforcement.",
        "evidence_times": [
          105.5,
          115.5,
          125.5
        ]
      },
      "codes": [
        {
          "code": "permission-step",
          "beat_indices": [
            3
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            0,
            4
          ]
        }
      ],
      "uncertainty": "The sampled sharing and approval states do not independently verify backend changes.",
      "media": {
        "sha256": "6decab1819d30a9be0397c82042079824096d78af96ca830a40019f4c0f5b1c7",
        "duration": 143.674,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:07.288023+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 387,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "6decab1819d30a9be0397c82042079824096d78af96ca830a40019f4c0f5b1c7"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "Is2NHa7awWY",
      "rank": 23,
      "title": "Build agent-ready sites with WebMCP",
      "source_url": "https://www.youtube.com/watch?v=Is2NHa7awWY",
      "publication_date": "2026-08-25",
      "publication_timestamp": "2026-08-25T13:15:30-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A presenter identifies slow browser interaction as a problem. A modeling example precedes advice about agent-facing tools, leaving broader possibilities open rather than proving speed improvements.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills plus unverified ASR; analytical windows, no playback."
      },
      "beats": [
        {
          "start": 0,
          "end": 28,
          "role": "hook",
          "summary": "Presenter proposes website tools as an alternative.",
          "visual": "Man on teal sofa beneath large windows.",
          "evidence_time": 20.5,
          "mechanics": "Wide introduction alternates with facial close-ups.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 1.74
          }
        },
        {
          "start": 28,
          "end": 45,
          "role": "proof",
          "summary": "A modeling demo illustrates the proposal.",
          "visual": "Dark request pane; white model; panda-like face.",
          "evidence_time": 40.5,
          "mechanics": "Later split-view samples contain more model detail.",
          "asr_anchor": {
            "word_index": 180,
            "source_time": 28.34
          }
        },
        {
          "start": 45,
          "end": 105,
          "role": "qualification",
          "summary": "Presenter recommends tool refinement and feedback.",
          "visual": "Sofa interview, typing hands, downward facial close-up.",
          "evidence_time": 85.5,
          "mechanics": "Practical advice returns to human framing.",
          "asr_anchor": {
            "word_index": 284,
            "source_time": 45
          }
        },
        {
          "start": 105,
          "end": 137.014,
          "role": "close",
          "summary": "Collaboration possibilities remain an invitation to experiment.",
          "visual": "Medium and close presenter views; logo.",
          "evidence_time": 130.5,
          "mechanics": "No second application replaces the closing speaker.",
          "asr_anchor": {
            "word_index": 736,
            "source_time": 105.42
          }
        }
      ],
      "editing_lesson": {
        "start": 28,
        "end": 45,
        "observation": "Text-only request precedes a split workspace and increasingly detailed model.",
        "inference": "Interpretation: visible refinement makes the proposal concrete.",
        "adaptation": "Capture your request, initial render and revised render with matching viewport.",
        "caution": "Stills cannot measure latency or establish individual tool calls.",
        "evidence_times": [
          30.5,
          35.5,
          40.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            2
          ]
        }
      ],
      "uncertainty": "Printability and interaction speed are presenter claims, not tested here.",
      "media": {
        "sha256": "b604b4968a346ab299ce25e82d137b2304db44d6a1b3cbab154dc45b8ff48b24",
        "duration": 137.014,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:08.008023+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 466,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "b604b4968a346ab299ce25e82d137b2304db44d6a1b3cbab154dc45b8ff48b24"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "2f9bBlGOuX4",
      "rank": 24,
      "title": "Codex, but make it weird. Hahahacks.",
      "source_url": "https://www.youtube.com/watch?v=2f9bBlGOuX4",
      "publication_date": "2026-08-25",
      "publication_timestamp": "2026-08-25T11:26:41-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "An event invites deliberately absurd inventions. Participant explanations and prototype glimpses develop that premise into a collective celebration, retaining an admitted limitation rather than presenting flawless products.",
      "family": "campaign",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; analytical windows, no continuous review."
      },
      "beats": [
        {
          "start": 0,
          "end": 20,
          "role": "hook",
          "summary": "Participants introduce playful hackathon ambitions.",
          "visual": "Wearable tubing, event room, floor-level laptop, food.",
          "evidence_time": 10.5,
          "mechanics": "Informal making scenes carry event identity.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 3.56
          }
        },
        {
          "start": 20,
          "end": 46,
          "role": "proof",
          "summary": "Makers describe shoe, mailbox and glasses experiments.",
          "visual": "Traffic light, colored laptop grid, improvised glasses.",
          "evidence_time": 40.5,
          "mechanics": "Distinct prototypes replace one another across samples.",
          "asr_anchor": {
            "word_index": 92,
            "source_time": 20.94
          }
        },
        {
          "start": 46,
          "end": 63,
          "role": "proof",
          "summary": "Another maker describes a water-based gag application.",
          "visual": "Tubing wearer, bottle, laptop code window.",
          "evidence_time": 55.5,
          "mechanics": "Physical apparatus precedes software and maker close-up.",
          "asr_anchor": {
            "word_index": 282,
            "source_time": 46.04
          }
        },
        {
          "start": 63,
          "end": 76.994,
          "role": "close",
          "summary": "Group appreciation concludes the experiments.",
          "visual": "Crowd, microphone speaker, white closing card.",
          "evidence_time": 70.5,
          "mechanics": "Individual demos yield to communal framing.",
          "asr_anchor": {
            "word_index": 380,
            "source_time": 65.319
          }
        }
      ],
      "editing_lesson": {
        "start": 35,
        "end": 46,
        "observation": "A maker indicates his eye, wears glasses, then smiles without them.",
        "inference": "Interpretation: the limitation keeps invention approachable.",
        "adaptation": "Film a maker explaining a constraint, trying the prototype and reporting its limitation.",
        "caution": "The eyesight benefit remains participant testimony, not medical evidence.",
        "evidence_times": [
          35.5,
          40.5,
          45.5
        ]
      },
      "codes": [
        {
          "code": "output-montage",
          "beat_indices": [
            1,
            2
          ]
        }
      ],
      "uncertainty": "Samples cannot establish the claimed shoe detection or water-trigger behavior.",
      "media": {
        "sha256": "848610e8bff93e6815f36579b75b0cdd1dad89fe0d01ddafe665e2526f044bdc",
        "duration": 76.994,
        "width": 854,
        "height": 480,
        "fps": "30000/1001",
        "retrieved_at": "2026-09-17T18:16:07.457333+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 210,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "848610e8bff93e6815f36579b75b0cdd1dad89fe0d01ddafe665e2526f044bdc"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "9MIkJHcMtVA",
      "rank": 25,
      "title": "No Limits on Text Chats with ChatGPT",
      "source_url": "https://www.youtube.com/watch?v=9MIkJHcMtVA",
      "publication_date": "2026-08-25",
      "publication_timestamp": "2026-08-25T10:21:58-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "The film associates chat availability with exuberant performance rather than a user task. Repeated promotional captions accompany performers and pinball inserts before branding closes, without demonstrating access conditions.",
      "family": "campaign",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills; ASR contains lyrics, not product narration; analytical windows, no listening."
      },
      "beats": [
        {
          "start": 0,
          "end": 30,
          "role": "context",
          "summary": "Establishes a playful performance setting.",
          "visual": "Striped set, performers, pinball hardware and digits.",
          "evidence_time": 20.5,
          "mechanics": "Broad set views precede isolated scoreboard detail.",
          "asr_anchor": null
        },
        {
          "start": 30,
          "end": 95,
          "role": "hook",
          "summary": "Captions promote unrestricted everyday text chatting.",
          "visual": "Female and male performers against colored geometry.",
          "evidence_time": 50.5,
          "mechanics": "The availability claim recurs over different poses.",
          "asr_anchor": null
        },
        {
          "start": 95,
          "end": 170,
          "role": "action",
          "summary": "Performance sustains the promotional association.",
          "visual": "Pinball playfield returns among performer close-ups.",
          "evidence_time": 105.5,
          "mechanics": "Mechanical inserts interrupt recurring human imagery.",
          "asr_anchor": null
        },
        {
          "start": 170,
          "end": 195.674,
          "role": "close",
          "summary": "Repeated claim resolves into product branding.",
          "visual": "Captioned performer; final illustrated pinball artwork and wordmark.",
          "evidence_time": 195.474,
          "mechanics": "Branding occupies the final sampled image.",
          "asr_anchor": null
        }
      ],
      "editing_lesson": {
        "start": 0,
        "end": 30,
        "observation": "Colorful spatial context precedes pinball hardware and enlarged score digits.",
        "inference": "Interpretation: scoring imagery supplies a playful counterpoint to availability.",
        "adaptation": "Establish your original setting, isolate its counter, then return to people.",
        "caution": "The scoreboard is scenery, not a usage meter or service proof.",
        "evidence_times": [
          0.5,
          15.5,
          20.5
        ]
      },
      "codes": [
        {
          "code": "context-before-detail",
          "beat_indices": [
            0
          ]
        }
      ],
      "uncertainty": "Spoken product narration is absent from ASR. Lyrics do not validate captions; musical synchronization is unreviewed.",
      "media": {
        "sha256": "a21e84cf79255cca8f32d2aa3b64656ba1b82354e659b6e63b683fa18d023840",
        "duration": 195.674,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:09.126552+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 332,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "a21e84cf79255cca8f32d2aa3b64656ba1b82354e659b6e63b683fa18d023840"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "jzmNh8lbSp8",
      "rank": 26,
      "title": "Make information visual with ChatGPT",
      "source_url": "https://www.youtube.com/watch?v=jzmNh8lbSp8",
      "publication_date": "2026-08-24",
      "publication_timestamp": "2026-08-24T16:09:56-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Dense textual information motivates a visual alternative. Meeting-note outputs develop into calendar, interaction and sharing examples before the presenter recommends trying the workflow.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; analytical windows, no playback."
      },
      "beats": [
        {
          "start": 0,
          "end": 32,
          "role": "context",
          "summary": "Presenter introduces visual handling of dense notes.",
          "visual": "Laptop opening; presenter at round table.",
          "evidence_time": 25.5,
          "mechanics": "Medium and close human views precede output.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 3.06
          }
        },
        {
          "start": 32,
          "end": 50,
          "role": "proof",
          "summary": "Meeting notes become structured planning views.",
          "visual": "Decision summary, timeline, calendar; presenter inset.",
          "evidence_time": 45.5,
          "mechanics": "Same example gains a different visual organization.",
          "asr_anchor": {
            "word_index": 190,
            "source_time": 32.08
          }
        },
        {
          "start": 50,
          "end": 68,
          "role": "action",
          "summary": "Examples extend to interaction and sharing.",
          "visual": "Tabs, contextual prompt button, export menu.",
          "evidence_time": 65.5,
          "mechanics": "Output controls replace the overview's broad framing.",
          "asr_anchor": {
            "word_index": 290,
            "source_time": 50.06
          }
        },
        {
          "start": 68,
          "end": 86.034,
          "role": "close",
          "summary": "Presenter offers personal endorsement and invitation.",
          "visual": "Full presenter returns, followed by logo.",
          "evidence_time": 75.5,
          "mechanics": "The screen inset gives way to interview framing.",
          "asr_anchor": {
            "word_index": 420,
            "source_time": 68.06
          }
        }
      ],
      "editing_lesson": {
        "start": 32,
        "end": 50,
        "observation": "Summary, detailed timeline and calendar share one meeting example.",
        "inference": "Interpretation: retained context makes iteration legible.",
        "adaptation": "Show your source summary, one requested reorganization and the resulting calendar.",
        "caution": "Different states do not establish the generation delay between them.",
        "evidence_times": [
          35.5,
          40.5,
          45.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            3
          ]
        }
      ],
      "uncertainty": "The visible publishing option does not establish a successfully shared site.",
      "media": {
        "sha256": "303b8997e996a31751ac2ef1b2da09d5461b7a518bc0e815b2cf45b4844548ad",
        "duration": 86.034,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:08.760494+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 259,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "303b8997e996a31751ac2ef1b2da09d5461b7a518bc0e815b2cf45b4844548ad"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "ceBruD6v5Bk",
      "rank": 27,
      "title": "You can just use your voice | ChatGPT Work",
      "source_url": "https://www.youtube.com/watch?v=ceBruD6v5Bk",
      "publication_date": "2026-08-24",
      "publication_timestamp": "2026-08-24T06:55:45-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A recipe revision prompts a spoken comparison request. Voice-interface states and a recipe document lead to an ingredient explanation, leaving requested notes and approval unresolved in samples.",
      "family": "product-film",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; analytical windows, no playback."
      },
      "beats": [
        {
          "start": 0,
          "end": 8,
          "role": "hook",
          "summary": "Voice controls introduce an ongoing recipe discussion.",
          "visual": "Logo followed by speaker, microphone and close controls.",
          "evidence_time": 5.5,
          "mechanics": "Brand field becomes enlarged interface detail.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 5.92
          }
        },
        {
          "start": 8,
          "end": 19,
          "role": "task",
          "summary": "User requests version comparison and compiled notes.",
          "visual": "Request bubble above purple voice orb.",
          "evidence_time": 15.5,
          "mechanics": "Later sample contains a longer request.",
          "asr_anchor": {
            "word_index": 10,
            "source_time": 8.02
          }
        },
        {
          "start": 19,
          "end": 29,
          "role": "action",
          "summary": "Assistant indicates comparison work.",
          "visual": "Processing status, then recipe document on crowded desktop.",
          "evidence_time": 25.5,
          "mechanics": "Isolated voice UI expands into document context.",
          "asr_anchor": {
            "word_index": 70,
            "source_time": 20.28
          }
        },
        {
          "start": 29,
          "end": 39.574,
          "role": "close",
          "summary": "Assistant attributes extra heat to one ingredient.",
          "visual": "Response text above orb; blue closing lettering.",
          "evidence_time": 30.5,
          "mechanics": "Desktop context yields to response and branding.",
          "asr_anchor": {
            "word_index": 96,
            "source_time": 30.28
          }
        }
      ],
      "editing_lesson": {
        "start": 8,
        "end": 34,
        "observation": "Request, recipe document and response occupy successive sampled layouts.",
        "inference": "Interpretation: document context connects instruction to explanation.",
        "adaptation": "Record your spoken task, relevant source document and attributable answer.",
        "caution": "Displayed processing duration is not this montage's elapsed time.",
        "evidence_times": [
          15.5,
          25.5,
          30.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            2,
            3
          ]
        }
      ],
      "uncertainty": "No sampled compiled notes or approval outcome; ingredient explanation is the film's claim.",
      "media": {
        "sha256": "0329b59cedd82a6b072900ed9d211a76d3b5d3b93a17785dcd2d5405a8702490",
        "duration": 39.574,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:16:08.415640+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 53,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "0329b59cedd82a6b072900ed9d211a76d3b5d3b93a17785dcd2d5405a8702490"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "XjSJ6ybS9I8",
      "rank": 28,
      "title": "How to Turn a Business Question Into a Strategy Deck With ChatGPT Work | Tutorial",
      "source_url": "https://www.youtube.com/watch?v=XjSJ6ybS9I8",
      "publication_date": "2026-08-18",
      "publication_timestamp": "2026-08-18T12:05:08-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A market-prioritization question is paired with an early deck preview. Numbered setup, source connection and storyline review explain the proposed route back to a finished readout.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; analytical windows, no playback."
      },
      "beats": [
        {
          "start": 0,
          "end": 33,
          "role": "hook",
          "summary": "Presenter previews a market recommendation deck.",
          "visual": "Green backdrop, slide insets, blue mascot.",
          "evidence_time": 15.5,
          "mechanics": "Output pages accompany the speaker before instructions.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 4.22
          }
        },
        {
          "start": 33,
          "end": 59,
          "role": "task",
          "summary": "Open workspace, pose question and connect sources.",
          "visual": "Laptop close-ups, numbered headings, enlarged prompts.",
          "evidence_time": 55.5,
          "mechanics": "Headings alternate with concrete request text.",
          "asr_anchor": {
            "word_index": 138,
            "source_time": 33.28
          }
        },
        {
          "start": 59,
          "end": 82,
          "role": "qualification",
          "summary": "Presenter calls for planning and source review.",
          "visual": "Storyline heading, deck-structure prompt, seated presenter.",
          "evidence_time": 65.5,
          "mechanics": "Planning prompt precedes later production request.",
          "asr_anchor": {
            "word_index": 262,
            "source_time": 59.06
          }
        },
        {
          "start": 82,
          "end": 107.914,
          "role": "close",
          "summary": "Final assembly returns to the readout promise.",
          "visual": "Production prompt, table slide inset, presenter, mascot endcard.",
          "evidence_time": 90.5,
          "mechanics": "Request is followed by another output glimpse.",
          "asr_anchor": {
            "word_index": 374,
            "source_time": 82.42
          }
        }
      ],
      "editing_lesson": {
        "start": 33,
        "end": 59,
        "observation": "Task heading, business prompt and source-specific prompt occupy distinct samples.",
        "inference": "Interpretation: headings make dependencies easier to retain.",
        "adaptation": "Name your step, show its exact input, then identify required sources.",
        "caution": "Prompt overlays do not demonstrate successful connection authorization.",
        "evidence_times": [
          40.5,
          45.5,
          55.5
        ]
      },
      "codes": [
        {
          "code": "question-led",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "code": "chapter-reset",
          "beat_indices": [
            1,
            2
          ]
        }
      ],
      "uncertainty": "Demonstration data and small slide insets do not validate the strategy or calculations.",
      "media": {
        "sha256": "c2b6e0a135012aedb6ead3b7b34ec412a532fddb8e61a79bd770f302025f8cb9",
        "duration": 107.914,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:09.520426+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 230,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "c2b6e0a135012aedb6ead3b7b34ec412a532fddb8e61a79bd770f302025f8cb9"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "0j9yDUDMrBs",
      "rank": 29,
      "title": "How to Create a Solid Blog Draft With ChatGPT Work | Tutorial",
      "source_url": "https://www.youtube.com/watch?v=0j9yDUDMrBs",
      "publication_date": "2026-08-18",
      "publication_timestamp": "2026-08-18T12:02:05-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Scattered launch materials create a writing bottleneck. Source gathering develops into template guidance, testimonial extraction and update requests, resolving in a claimed shareable draft rather than an inspected publication.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; analytical windows, no playback."
      },
      "beats": [
        {
          "start": 0,
          "end": 47,
          "role": "context",
          "summary": "Presenter describes assembling launch information manually.",
          "visual": "Purple backdrop, laptop, mascot, floating source-app icons.",
          "evidence_time": 30.5,
          "mechanics": "Scattered icons visualize the described source burden.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 4.14
          }
        },
        {
          "start": 47,
          "end": 78,
          "role": "task",
          "summary": "Gather sources and dictate an initial request.",
          "visual": "Numbered setup cards, source prompt, waveform overlay.",
          "evidence_time": 75.5,
          "mechanics": "Keyboard framing later contrasts with voice graphics.",
          "asr_anchor": {
            "word_index": 258,
            "source_time": 47.14
          }
        },
        {
          "start": 78,
          "end": 103,
          "role": "action",
          "summary": "Supply format guidance before enriching the draft.",
          "visual": "Template prompt, side-angle and centered presenter.",
          "evidence_time": 80.5,
          "mechanics": "Specific input gives way to explanation.",
          "asr_anchor": {
            "word_index": 448,
            "source_time": 78.16
          }
        },
        {
          "start": 103,
          "end": 134,
          "role": "action",
          "summary": "Use interview sources for customer material.",
          "visual": "Testimonial panel, file list, transcript-selection prompt.",
          "evidence_time": 115.5,
          "mechanics": "Example panel precedes source folder and extraction request.",
          "asr_anchor": {
            "word_index": 610,
            "source_time": 103.04
          }
        },
        {
          "start": 134,
          "end": 152.594,
          "role": "close",
          "summary": "Request monitoring and recap the claimed draft outcome.",
          "visual": "Scheduled-task prompt, presenter, mascot endcard.",
          "evidence_time": 135.5,
          "mechanics": "Final instruction precedes human summary.",
          "asr_anchor": {
            "word_index": 806,
            "source_time": 134.18
          }
        }
      ],
      "editing_lesson": {
        "start": 103,
        "end": 134,
        "observation": "A testimonial example precedes interview files and an extraction prompt.",
        "inference": "Interpretation: desired output gives source selection a purpose.",
        "adaptation": "Preview your intended component, show authorized sources, then specify extraction constraints.",
        "caution": "The preview does not establish faithful quotation or speaker consent.",
        "evidence_times": [
          105.5,
          115.5,
          130.5
        ]
      },
      "codes": [
        {
          "code": "context-before-detail",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "code": "chapter-reset",
          "beat_indices": [
            1
          ]
        }
      ],
      "uncertainty": "No complete published draft or executed monitoring cycle is visible in samples.",
      "media": {
        "sha256": "0a2b0b537707b5fee4df6d61d3fed6f4daf1635b2e208e92e3266ca2747db9ab",
        "duration": 152.594,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:09.537318+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 447,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "0a2b0b537707b5fee4df6d61d3fed6f4daf1635b2e208e92e3266ca2747db9ab"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "yQZgOSHHxjk",
      "rank": 30,
      "title": "How to Prepare for Customer Meetings With ChatGPT Work | Tutorial",
      "source_url": "https://www.youtube.com/watch?v=yQZgOSHHxjk",
      "publication_date": "2026-08-18",
      "publication_timestamp": "2026-08-18T11:58:56-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Meeting preparation and follow-up compete with customer attention. Output previews lead into connected sources, mobile requests and approval instructions, ending with a brief and a renewed customer-focus claim.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; analytical windows, no playback."
      },
      "beats": [
        {
          "start": 0,
          "end": 33,
          "role": "hook",
          "summary": "Presenter contrasts administrative work with customer time.",
          "visual": "Green backdrop, document previews, orange mascot.",
          "evidence_time": 10.5,
          "mechanics": "Output insets appear before numbered instructions.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 4.14
          }
        },
        {
          "start": 33,
          "end": 52,
          "role": "task",
          "summary": "Connect tools and specify meeting context.",
          "visual": "Numbered headings and laptop prompt overlays.",
          "evidence_time": 50.5,
          "mechanics": "Connection request precedes meeting-specific request.",
          "asr_anchor": {
            "word_index": 172,
            "source_time": 33.8
          }
        },
        {
          "start": 52,
          "end": 69,
          "role": "action",
          "summary": "Presenter describes requesting briefs while traveling.",
          "visual": "Seated presenter followed by phone and waveform.",
          "evidence_time": 65.5,
          "mechanics": "Phone close-up substitutes for desktop framing.",
          "asr_anchor": {
            "word_index": 274,
            "source_time": 52.18
          }
        },
        {
          "start": 69,
          "end": 83,
          "role": "qualification",
          "summary": "Delegate follow-up while retaining approval responsibility.",
          "visual": "Follow-up heading, request overlay, approval heading.",
          "evidence_time": 80.5,
          "mechanics": "Delegation and review receive separate textual emphasis.",
          "asr_anchor": {
            "word_index": 384,
            "source_time": 69.12
          }
        },
        {
          "start": 83,
          "end": 100.214,
          "role": "close",
          "summary": "A brief supports the concluding customer-focus claim.",
          "visual": "Document inset, presenter, final logo.",
          "evidence_time": 85.5,
          "mechanics": "Artifact glimpse precedes closing human framing.",
          "asr_anchor": {
            "word_index": 460,
            "source_time": 83.56
          }
        }
      ],
      "editing_lesson": {
        "start": 69,
        "end": 83,
        "observation": "Follow-up heading precedes a concrete request and separate approval heading.",
        "inference": "Interpretation: review remains visible after delegation.",
        "adaptation": "Show your requested follow-up, draft result and explicit human approval checkpoint.",
        "caution": "The heading is instructional; no approval transaction is demonstrated here.",
        "evidence_times": [
          70.5,
          75.5,
          80.5
        ]
      },
      "codes": [
        {
          "code": "chapter-reset",
          "beat_indices": [
            1,
            3
          ]
        }
      ],
      "uncertainty": "Proactive preparation and time savings remain presenter claims; samples do not establish scheduling behavior.",
      "media": {
        "sha256": "e224f881bc848d7b4f8e3bd28182be03e1f6a2a156d5592267d28a47f7eeea0f",
        "duration": 100.214,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:10.443627+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 264,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "e224f881bc848d7b4f8e3bd28182be03e1f6a2a156d5592267d28a47f7eeea0f"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "wIf9yMGalPU",
      "rank": 31,
      "title": "You can just keep the work moving | ChatGPT Work",
      "source_url": "https://www.youtube.com/watch?v=wIf9yMGalPU",
      "publication_date": "2026-08-17",
      "publication_timestamp": "2026-08-17T11:25:48-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "An overnight monitoring request sets up work continuing after sign-off. Desktop activity and a populated findings report provide a visible payoff before a meeting-linked follow-up extends the delegation.",
      "family": "product-film",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills plus source checks at 27.5, 28.5 and 29.5 seconds, with event-only ASR; no continuous playback; analytical windows, not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 15,
          "role": "task",
          "summary": "Delegate overnight launch monitoring through a prompt.",
          "visual": "Logo, enlarged composer, then a populated request with connectors.",
          "evidence_time": 10.5,
          "mechanics": "Later composer sample contains substantially more task context.",
          "asr_anchor": null
        },
        {
          "start": 15,
          "end": 30,
          "role": "proof",
          "summary": "Checking activity leads to a populated report with ranked findings.",
          "visual": "Desktop Slack context precedes source coverage and three ranked findings with follow-up owners.",
          "evidence_time": 29.5,
          "mechanics": "The report overview is followed by more specific source and findings details.",
          "asr_anchor": null
        },
        {
          "start": 30,
          "end": 40,
          "role": "task",
          "summary": "A follow-up requests recurring meeting-linked updates.",
          "visual": "Large composer fragment precedes a right-aligned message bubble.",
          "evidence_time": 35.5,
          "mechanics": "Entry and submitted-looking message are separate sampled states.",
          "asr_anchor": null
        },
        {
          "start": 40,
          "end": 45.734,
          "role": "close",
          "summary": "A continuity promise replaces the workflow.",
          "visual": "Centered two-line slogan, then black knot on white.",
          "evidence_time": 40.5,
          "mechanics": "Text and symbol occupy otherwise empty closing frames.",
          "asr_anchor": null
        }
      ],
      "editing_lesson": {
        "start": 5,
        "end": 40,
        "observation": "A monitoring prompt precedes a populated findings report, followed by a new meeting-linked request.",
        "inference": "Interpretation: the displayed payoff gives the next delegation a reason to follow.",
        "adaptation": "Show your monitoring request, hold its actual findings long enough to read, then make a follow-up grounded in that result.",
        "caution": "A displayed report does not independently verify completed monitoring, accuracy or future recurring execution.",
        "evidence_times": [
          10.5,
          29.5,
          35.5
        ]
      },
      "codes": [
        {
          "code": "context-before-detail",
          "beat_indices": [
            0,
            1
          ]
        }
      ],
      "uncertainty": "Speech evidence absent; ASR labels an audio event, not validated music listening.",
      "media": {
        "sha256": "d245b10e6e881e79ecc47eebfa977ab574dd44155f3bc8231bebc5037b304f2c",
        "duration": 45.734,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:16:10.040220+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 0,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "d245b10e6e881e79ecc47eebfa977ab574dd44155f3bc8231bebc5037b304f2c"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "Cxp2aZQqkYM",
      "rank": 32,
      "title": "How Base44 Uses GPT-5.6 to Build Apps With 20% Fewer Tokens",
      "source_url": "https://www.youtube.com/watch?v=Cxp2aZQqkYM",
      "publication_date": "2026-08-17",
      "publication_timestamp": "2026-08-17T08:22:51-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A speaker presents better results with fewer tokens as the problem's answer. Builder screens give that claim concrete examples before testing assertions and customer delivery return the story to testimony.",
      "family": "testimonial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; no playback; analytical windows, not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 16,
          "role": "hook",
          "summary": "Speaker claims comparative quality and token savings.",
          "visual": "Seated man against cubbies; large identification graphic appears.",
          "evidence_time": 10.5,
          "mechanics": "Tighter face samples precede wider framing with oversized type.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.08
          }
        },
        {
          "start": 16,
          "end": 43,
          "role": "proof",
          "summary": "App examples illustrate the speaker's building claims.",
          "visual": "Builder composer, coastal map, and mobile weather interface.",
          "evidence_time": 30.5,
          "mechanics": "Full builder views alternate with enlarged prompt-and-preview crops.",
          "asr_anchor": {
            "word_index": 64,
            "source_time": 16.239
          }
        },
        {
          "start": 43,
          "end": 55,
          "role": "qualification",
          "summary": "Speaker describes testing before production deployment.",
          "visual": "Same seated speaker fills both sampled frames.",
          "evidence_time": 45.5,
          "mechanics": "Product examples yield to direct testimony, without benchmark tables.",
          "asr_anchor": {
            "word_index": 196,
            "source_time": 43.92
          }
        },
        {
          "start": 55,
          "end": 71.134,
          "role": "close",
          "summary": "Less correction is framed as faster customer delivery.",
          "visual": "Map and follow-up composer, returning speaker, final knot.",
          "evidence_time": 60.5,
          "mechanics": "Concrete refinement request precedes human conclusion and blank-field branding.",
          "asr_anchor": {
            "word_index": 260,
            "source_time": 55.06
          }
        }
      ],
      "editing_lesson": {
        "start": 20,
        "end": 43,
        "observation": "Builder context precedes map and weather previews.",
        "inference": "Interpretation: artifact variety makes the platform claim tangible.",
        "adaptation": "Show your request, resulting application, then a separate clearly identified example.",
        "caution": "Samples neither establish generation latency nor verify token savings.",
        "evidence_times": [
          20.5,
          30.5,
          40.5
        ]
      },
      "codes": [
        {
          "code": "presenter-return",
          "beat_indices": [
            0,
            1,
            2
          ]
        },
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1
          ]
        }
      ],
      "uncertainty": "Comparative performance remains the speaker's claim; testing methodology is not shown.",
      "media": {
        "sha256": "b1a2b50f5662e9d8607b9d4e9426169829a703e6d0d7639e3ac78a10c7d9106d",
        "duration": 71.134,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:10.595738+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 166,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "b1a2b50f5662e9d8607b9d4e9426169829a703e6d0d7639e3ac78a10c7d9106d"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "EDWbX_-RYws",
      "rank": 33,
      "title": "Previewing Ultrafast mode: GPT‑5.6 Sol at up to 14X the speed",
      "source_url": "https://www.youtube.com/watch?v=EDWbX_-RYws",
      "publication_date": "2026-08-14",
      "publication_timestamp": "2026-08-14T10:19:17-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Urgent outages and interrupted concentration make speed a workplace problem. Staff accounts alternate with incident-analysis screens, resolving into claimed continuity of work rather than a measured comparison.",
      "family": "testimonial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; no playback; analytical windows, not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 14,
          "role": "hook",
          "summary": "Staff introduce exceptional speed and outage urgency.",
          "visual": "Three webcam speakers; oversized mode title over one.",
          "evidence_time": 10.5,
          "mechanics": "Different rooms and lower thirds establish multiple witnesses.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.12
          }
        },
        {
          "start": 14,
          "end": 28,
          "role": "task",
          "summary": "Incident investigation becomes a concrete agent request.",
          "visual": "Prompt crop, full application, then green-shirted speaker.",
          "evidence_time": 15.5,
          "mechanics": "Small speaker inset accompanies the otherwise dominant interface.",
          "asr_anchor": {
            "word_index": 76,
            "source_time": 14.2
          }
        },
        {
          "start": 28,
          "end": 46,
          "role": "proof",
          "summary": "A speaker claims shorter investigation preparation.",
          "visual": "Black-shirted speaker, application overview, enlarged incident timeline.",
          "evidence_time": 40.5,
          "mechanics": "Timeline crop makes evidence lines larger than overview text.",
          "asr_anchor": {
            "word_index": 176,
            "source_time": 28.5
          }
        },
        {
          "start": 46,
          "end": 64,
          "role": "proof",
          "summary": "Parallel searching supports claims of fewer attention breaks.",
          "visual": "Investigation prompt followed by a three-part diagnosis response.",
          "evidence_time": 60.5,
          "mechanics": "Prompt and response crops retain different speakers as insets.",
          "asr_anchor": {
            "word_index": 278,
            "source_time": 46.18
          }
        },
        {
          "start": 64,
          "end": 92.974,
          "role": "close",
          "summary": "Staff associate speed with sustained coding momentum.",
          "visual": "Several faces, one application inset layout, final wordmark.",
          "evidence_time": 85.5,
          "mechanics": "Full-face testimony dominates after the detailed response samples.",
          "asr_anchor": {
            "word_index": 384,
            "source_time": 64.04
          }
        }
      ],
      "editing_lesson": {
        "start": 28,
        "end": 46,
        "observation": "Speaker, application overview, and enlarged timeline form successive sampled states.",
        "inference": "Interpretation: increasing specificity gives testimony an inspectable referent.",
        "adaptation": "Introduce a real incident, show its workspace, enlarge the diagnostic result.",
        "caution": "Displayed results do not validate claimed elapsed-time improvements.",
        "evidence_times": [
          30.5,
          35.5,
          40.5
        ]
      },
      "codes": [
        {
          "code": "human-stakes",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            1,
            2,
            4
          ]
        }
      ],
      "uncertainty": "No continuous timing test; concurrency and quality are attributed staff assertions.",
      "media": {
        "sha256": "4163509298e889aa648044b183d1241b843929fddbf82ce80f07efe6232efd32",
        "duration": 92.974,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:11.015235+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 280,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "4163509298e889aa648044b183d1241b843929fddbf82ce80f07efe6232efd32"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "0XPk_MAwCW4",
      "rank": 34,
      "title": "Inside Cricket’s Smartest Backroom | Rajasthan Royals | ChatGPT @rajasthanroyals",
      "source_url": "https://www.youtube.com/watch?v=0XPk_MAwCW4",
      "publication_date": "2026-08-13",
      "publication_timestamp": "2026-08-13T23:30:28-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A joking reluctance to reveal team secrets introduces preparation behind cricket performance. Auction, creative, ticketing, and student examples broaden the account before the opening coach returns for an informal conclusion.",
      "family": "testimonial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; no playback; analytical windows, not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 24,
          "role": "hook",
          "summary": "Off-field preparation is framed as consequential.",
          "visual": "Coach on sofa, doorway angle, outdoor batting practice.",
          "evidence_time": 20.5,
          "mechanics": "Formal interview and practice imagery share the opening window.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 0.62
          }
        },
        {
          "start": 24,
          "end": 68,
          "role": "context",
          "summary": "Staff describe auction complexity and direct statistical queries.",
          "visual": "Alternating staff interviews, player-card overlay, extreme facial close-up.",
          "evidence_time": 35.5,
          "mechanics": "Player portraits accompany discussion; query interface is not sampled.",
          "asr_anchor": {
            "word_index": 142,
            "source_time": 24.6
          }
        },
        {
          "start": 68,
          "end": 95,
          "role": "proof",
          "summary": "Creative staff describe an AI-assisted heritage campaign.",
          "visual": "Two interviewees flank a three-panel player-image composition.",
          "evidence_time": 85.5,
          "mechanics": "Vertical campaign examples replace the room-bound interview framing.",
          "asr_anchor": {
            "word_index": 470,
            "source_time": 68.62
          }
        },
        {
          "start": 95,
          "end": 123,
          "role": "discussion",
          "summary": "Organizational adoption expands into ticketing recommendations.",
          "visual": "Interview with line-chart and circular-metric overlays; laptop hands.",
          "evidence_time": 110.5,
          "mechanics": "Data graphics sit beside the speaker rather than replacing him.",
          "asr_anchor": {
            "word_index": 636,
            "source_time": 95.24
          }
        },
        {
          "start": 123,
          "end": 173,
          "role": "proof",
          "summary": "Student strategies reportedly influence auction decisions.",
          "visual": "Interview, student laptop groups, and glass-walled presentation room.",
          "evidence_time": 155.5,
          "mechanics": "Participants and meeting footage interrupt the recurring seated account.",
          "asr_anchor": {
            "word_index": 812,
            "source_time": 123.66
          }
        },
        {
          "start": 173,
          "end": 214.734,
          "role": "close",
          "summary": "Future possibilities yield to the coach's closing joke.",
          "visual": "Staff close-ups, original sofa setting, crew presence, partial wordmark.",
          "evidence_time": 205.5,
          "mechanics": "Opening interview subject returns with less formal surrounding activity.",
          "asr_anchor": {
            "word_index": 1156,
            "source_time": 173.44
          }
        }
      ],
      "editing_lesson": {
        "start": 123,
        "end": 163,
        "observation": "Student collaboration and a later presentation accompany the recruitment account.",
        "inference": "Interpretation: named participation becomes a visible human process.",
        "adaptation": "Pair an original participant account with their work session and decision meeting.",
        "caution": "Meeting footage cannot prove which purchases the strategy caused.",
        "evidence_times": [
          135.5,
          145.5,
          155.5
        ]
      },
      "codes": [
        {
          "code": "presenter-return",
          "beat_indices": [
            0,
            5
          ]
        }
      ],
      "uncertainty": "Campaign production savings and auction effects remain attributed claims, not demonstrated outcomes.",
      "media": {
        "sha256": "d50c6de5a19aba7b2d17f5a2b15d23c179326547bbd937820dfc11e8bed559d3",
        "duration": 214.734,
        "width": 854,
        "height": 480,
        "fps": "25/1",
        "retrieved_at": "2026-09-17T18:16:12.754509+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 682,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "d50c6de5a19aba7b2d17f5a2b15d23c179326547bbd937820dfc11e8bed559d3"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "W-HhMUe9hOg",
      "rank": 35,
      "title": "Computer History in ChatGPT",
      "source_url": "https://www.youtube.com/watch?v=W-HhMUe9hOg",
      "publication_date": "2026-08-13",
      "publication_timestamp": "2026-08-13T13:34:04-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "The presenter asks how an agent can understand work without repeated context. Retrieval, sharing, and recap examples precede controls and suggested reuse, making personalization conditional on inspectable history.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; no playback; analytical windows, not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 23,
          "role": "context",
          "summary": "Presenter introduces enabling activity-based memory.",
          "visual": "Standing presenter behind laptop, then angled facial close-ups.",
          "evidence_time": 5.5,
          "mechanics": "Wider desk context precedes sustained close framing.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.24
          }
        },
        {
          "start": 23,
          "end": 68,
          "role": "action",
          "summary": "Document retrieval leads to a sharing request.",
          "visual": "Chat responses, purple orb, intermittent presenter inset.",
          "evidence_time": 65.5,
          "mechanics": "Later samples preserve earlier turns above added sharing confirmation.",
          "asr_anchor": {
            "word_index": 154,
            "source_time": 24.44
          }
        },
        {
          "start": 68,
          "end": 99,
          "role": "proof",
          "summary": "A morning recap expands the remembered context.",
          "visual": "Longer conversation fills the screen; presenter returns.",
          "evidence_time": 90.5,
          "mechanics": "Recap text accumulates beneath the prior document-sharing exchange.",
          "asr_anchor": {
            "word_index": 412,
            "source_time": 68.08
          }
        },
        {
          "start": 99,
          "end": 147,
          "role": "qualification",
          "summary": "Presenter explains capture limits, exclusions, and deletion.",
          "visual": "Presenter, exclusion dialog, and detailed activity-history entries.",
          "evidence_time": 125.5,
          "mechanics": "Settings and history views replace conversational output examples.",
          "asr_anchor": {
            "word_index": 566,
            "source_time": 99.16
          }
        },
        {
          "start": 147,
          "end": 190.214,
          "role": "close",
          "summary": "Suggested skills extend history into repeatable work.",
          "visual": "History list, populated skill prompt, presenter, final knot.",
          "evidence_time": 155.5,
          "mechanics": "History-derived suggestion becomes composer content before the presenter concludes.",
          "asr_anchor": {
            "word_index": 832,
            "source_time": 147.12
          }
        }
      ],
      "editing_lesson": {
        "start": 115,
        "end": 156,
        "observation": "Exclusion settings precede inspectable history and a suggested-skill prompt.",
        "inference": "Interpretation: controls qualify the benefit before extending its scope.",
        "adaptation": "Show your actual exclusion control, retained evidence, then a permitted reuse.",
        "caution": "No sampled deletion result establishes that removal succeeded.",
        "evidence_times": [
          125.5,
          135.5,
          155.5
        ]
      },
      "codes": [
        {
          "code": "permission-step",
          "beat_indices": [
            0,
            3
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            1,
            2,
            4
          ]
        }
      ],
      "uncertainty": "Capture architecture and successful sharing are film assertions, not independently tested behavior.",
      "media": {
        "sha256": "a8636b614d33741e30c83359b909ae6c0ac1714dcc5c37d9b85f9e068dcee7dc",
        "duration": 190.214,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:14.633164+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 542,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "a8636b614d33741e30c83359b909ae6c0ac1714dcc5c37d9b85f9e068dcee7dc"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "bl2eGejkbaU",
      "rank": 36,
      "title": "Get a daily CFO briefing with ChatGPT Work",
      "source_url": "https://www.youtube.com/watch?v=bl2eGejkbaU",
      "publication_date": "2026-08-13",
      "publication_timestamp": "2026-08-13T06:06:22-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Scattered operating and market signals create a prioritization problem for a finance leader. A phone briefing leads to desktop acquisition analysis and a scenario chart, presenting a path from noticing to testing.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; no playback; analytical windows, not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 28,
          "role": "context",
          "summary": "A mobile briefing organizes close-related priorities.",
          "visual": "White title cards; narrow phone dashboard against stars.",
          "evidence_time": 20.5,
          "mechanics": "Phone overview and enlarged checklist expose different information levels.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.24
          }
        },
        {
          "start": 28,
          "end": 58,
          "role": "task",
          "summary": "Exceptions and market signals prompt acquisition analysis.",
          "visual": "Contract cards, market cards, then mobile chat request.",
          "evidence_time": 55.5,
          "mechanics": "Same narrow device framing connects signal review to instruction.",
          "asr_anchor": {
            "word_index": 162,
            "source_time": 28.04
          }
        },
        {
          "start": 58,
          "end": 81,
          "role": "proof",
          "summary": "Memo and workbook expose the proposed acquisition case.",
          "visual": "Overlapping phone/desktop views, document pages, then Excel model.",
          "evidence_time": 65.5,
          "mechanics": "Desktop artifacts occupy progressively more of the star-backed field.",
          "asr_anchor": {
            "word_index": 342,
            "source_time": 58.08
          }
        },
        {
          "start": 81,
          "end": 108.294,
          "role": "close",
          "summary": "Scenario output precedes a decision-making benefit claim.",
          "visual": "Spreadsheet assistant pane, forecast chart, slogan, illustrated branding.",
          "evidence_time": 95.5,
          "mechanics": "Chart appears after model/request states and before closing cards.",
          "asr_anchor": {
            "word_index": 452,
            "source_time": 81.1
          }
        }
      ],
      "editing_lesson": {
        "start": 45,
        "end": 76,
        "observation": "Mobile signal, overlapping devices, and desktop workbook share one backdrop.",
        "inference": "Interpretation: device continuity connects discovery with deeper analysis.",
        "adaptation": "Carry one original issue from mobile request into its desktop result.",
        "caution": "The commute exists in narration, not location footage.",
        "evidence_times": [
          50.5,
          60.5,
          75.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "code": "proof-before-close",
          "beat_indices": [
            3
          ]
        }
      ],
      "uncertainty": "Demo forecasts and acquisition recommendations are illustrative film outputs, not validated financial conclusions.",
      "media": {
        "sha256": "710e3c2d2617986ca4860828f206141468cf000cdd54844e6ef610a88e282630",
        "duration": 108.294,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:16:12.421191+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 291,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "710e3c2d2617986ca4860828f206141468cf000cdd54844e6ef610a88e282630"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "94Td-y5z5rk",
      "rank": 37,
      "title": "Build custom financial forecasting apps with ChatGPT Work",
      "source_url": "https://www.youtube.com/watch?v=94Td-y5z5rk",
      "publication_date": "2026-08-13",
      "publication_timestamp": "2026-08-13T06:05:30-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A completed close still needs an understandable account of what comes next. The walkthrough develops a site into scenarios, supporting details, and board material before presenting a specific operating-change hypothesis.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; no playback; analytical windows, not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 24,
          "role": "task",
          "summary": "Request an interactive explanation of financial results.",
          "visual": "Title cards, blank composer, populated prompt, response window.",
          "evidence_time": 15.5,
          "mechanics": "White typography gives way to a star-backed application frame.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.24
          }
        },
        {
          "start": 24,
          "end": 48,
          "role": "proof",
          "summary": "Generated site organizes outcomes, scenarios, and account details.",
          "visual": "Preview cards, dashboard bars, scenario sliders, reconciliation table.",
          "evidence_time": 35.5,
          "mechanics": "Different site sections retain chat alongside the artifact.",
          "asr_anchor": {
            "word_index": 136,
            "source_time": 25.22
          }
        },
        {
          "start": 48,
          "end": 69,
          "role": "action",
          "summary": "Board material leads into a paired operating-change scenario.",
          "visual": "Four-part board page, dark-green cover, two assumption sliders.",
          "evidence_time": 65.5,
          "mechanics": "Later view exposes both sliders beneath unchanged headline metrics.",
          "asr_anchor": {
            "word_index": 252,
            "source_time": 48.14
          }
        },
        {
          "start": 69,
          "end": 83.154,
          "role": "close",
          "summary": "Cards generalize the scenario into planning value.",
          "visual": "Two centered statements followed by illustrated wordmark surround.",
          "evidence_time": 75.5,
          "mechanics": "Dense application disappears; sparse typography carries the conclusion.",
          "asr_anchor": {
            "word_index": 380,
            "source_time": 70.14
          }
        }
      ],
      "editing_lesson": {
        "start": 48,
        "end": 69,
        "observation": "Board framing precedes visible expense and partner-contribution controls.",
        "inference": "Interpretation: an abstract strategy becomes inspectable assumptions.",
        "adaptation": "Introduce your hypothesis, reveal its controls, then show a verified changed outcome.",
        "caution": "The inspected slider samples retain identical values; no recalculation is established.",
        "evidence_times": [
          50.5,
          60.5,
          65.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            0,
            1
          ]
        }
      ],
      "uncertainty": "Hidden calculations and forecast correctness cannot be inferred from visible site polish.",
      "media": {
        "sha256": "56d5a6ee862427271b8eaca9577460e217f98f0bfe9c948dfd210dd063990bbf",
        "duration": 83.154,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:16:13.577067+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 212,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "56d5a6ee862427271b8eaca9577460e217f98f0bfe9c948dfd210dd063990bbf"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "Ljm3nHQQiSk",
      "rank": 38,
      "title": "Reconcile quarter-end financials with ChatGPT Work",
      "source_url": "https://www.youtube.com/watch?v=Ljm3nHQQiSk",
      "publication_date": "2026-08-13",
      "publication_timestamp": "2026-08-13T06:04:31-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "The narrator asks whether a quarter can be closed defensibly. A conditional assessment exposes unresolved gates and supporting figures, ending with team instructions rather than demonstrated final clearance.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; no playback; analytical windows, not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 24,
          "role": "task",
          "summary": "A close-confidence question initiates cross-system reconciliation.",
          "visual": "Title cards, empty composer, enlarged request, response window.",
          "evidence_time": 15.5,
          "mechanics": "Request crop enlarges context before the report appears.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.38
          }
        },
        {
          "start": 24,
          "end": 40,
          "role": "qualification",
          "summary": "Assessment leaves four gates and residual risk.",
          "visual": "Close report; isolated status tile on darkened surroundings.",
          "evidence_time": 30.5,
          "mechanics": "Spotlight sample prioritizes conditional status over nearby figures.",
          "asr_anchor": {
            "word_index": 138,
            "source_time": 24.82
          }
        },
        {
          "start": 40,
          "end": 59,
          "role": "proof",
          "summary": "Report details explain unfavorable variances and posted actuals.",
          "visual": "Green/red waterfall, blue horizontal bars, account table.",
          "evidence_time": 50.5,
          "mechanics": "Aggregate bridge precedes ranked accounts and tabular records.",
          "asr_anchor": {
            "word_index": 224,
            "source_time": 40.52
          }
        },
        {
          "start": 59,
          "end": 90.214,
          "role": "close",
          "summary": "Outstanding gates become assigned team actions.",
          "visual": "Release-gate table, Teams message, closing statements and knot.",
          "evidence_time": 75.5,
          "mechanics": "Analytical report gives way to a communication surface before branding.",
          "asr_anchor": {
            "word_index": 322,
            "source_time": 59.16
          }
        }
      ],
      "editing_lesson": {
        "start": 24,
        "end": 59,
        "observation": "Isolated conditional status precedes variance chart and account evidence.",
        "inference": "Interpretation: the verdict gives subsequent detail a reason to matter.",
        "adaptation": "Show your real unresolved status, then trace it to supporting records.",
        "caution": "Do not replace conditional clearance with an unearned success ending.",
        "evidence_times": [
          30.5,
          45.5,
          55.5
        ]
      },
      "codes": [
        {
          "code": "question-led",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "code": "context-before-detail",
          "beat_indices": [
            1,
            2
          ]
        }
      ],
      "uncertainty": "No observed final gate release or independently verified accounting reconciliation.",
      "media": {
        "sha256": "7d5716d4d2d604cc15b64102bf57b1976b904650ea8df03fb8b69bd4f02d42b1",
        "duration": 90.214,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:16:12.876998+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 244,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "7d5716d4d2d604cc15b64102bf57b1976b904650ea8df03fb8b69bd4f02d42b1"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "_HCks5jkPLw",
      "rank": 39,
      "title": "Use ChatGPT Work to deliver board-ready reporting",
      "source_url": "https://www.youtube.com/watch?v=_HCks5jkPLw",
      "publication_date": "2026-08-13",
      "publication_timestamp": "2026-08-13T04:31:36-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Conflicting finance materials threaten publication readiness. A blocked assessment leads into source documents and marked discrepancies, while the presenter proposes rerunning checks without showing a cleared result.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; no playback; analytical windows, not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 39,
          "role": "context",
          "summary": "Presenter explains gathering materials and cross-checking systems.",
          "visual": "Title, application composer, assessment; small upper-left presenter tile.",
          "evidence_time": 20.5,
          "mechanics": "Assessment is already visible during narrated input explanation.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 3.16
          }
        },
        {
          "start": 39,
          "end": 73,
          "role": "qualification",
          "summary": "Two inconsistencies prevent publication readiness.",
          "visual": "Assessment tables and issue list within the same application.",
          "evidence_time": 65.5,
          "mechanics": "Later samples expose lower issue details beneath the verdict.",
          "asr_anchor": {
            "word_index": 238,
            "source_time": 39.02
          }
        },
        {
          "start": 73,
          "end": 111,
          "role": "action",
          "summary": "Source views locate corrections for the team.",
          "visual": "Spreadsheet, slide deck, and memo with red passages.",
          "evidence_time": 90.5,
          "mechanics": "Chat remains left while different source artifacts occupy right.",
          "asr_anchor": {
            "word_index": 464,
            "source_time": 73.0
          }
        },
        {
          "start": 111,
          "end": 150.554,
          "role": "close",
          "summary": "Presenter proposes repeated checks and broader document comparison.",
          "visual": "Marked memo, slide deck, full-size presenter, closing knot.",
          "evidence_time": 140.5,
          "mechanics": "Source inspection yields to the presenter without a cleared assessment.",
          "asr_anchor": {
            "word_index": 736,
            "source_time": 111.6
          }
        }
      ],
      "editing_lesson": {
        "start": 65,
        "end": 101,
        "observation": "Issue list precedes side-by-side spreadsheet and marked memo views.",
        "inference": "Interpretation: persistent chat preserves the diagnosis while locating evidence.",
        "adaptation": "Keep an original discrepancy visible while opening its affected source documents.",
        "caution": "Red markings do not establish that corrections were saved.",
        "evidence_times": [
          70.5,
          75.5,
          90.5
        ]
      },
      "codes": [
        {
          "code": "presenter-return",
          "beat_indices": [
            0,
            2,
            3
          ]
        }
      ],
      "uncertainty": "ASR contains disfluencies; neither a completed rerun nor successful publication is sampled.",
      "media": {
        "sha256": "688c21ded43405053b7ca91e7f048fd7304fce5fe379a83e498e39e78f83a609",
        "duration": 150.554,
        "width": 854,
        "height": 480,
        "fps": "24/1",
        "retrieved_at": "2026-09-17T18:16:13.696053+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 503,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "688c21ded43405053b7ca91e7f048fd7304fce5fe379a83e498e39e78f83a609"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "mOAHWk24D28",
      "rank": 40,
      "title": "Use ChatGPT Work to build custom forecasting apps",
      "source_url": "https://www.youtube.com/watch?v=mOAHWk24D28",
      "publication_date": "2026-08-13",
      "publication_timestamp": "2026-08-13T04:27:26-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A forecast preview precedes the presenter's account of monthly coordination pressure. The demonstration rebuilds context through a request, aggregate results, and account adjustments, leaving model-wide recalculation chiefly in narration.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second stills and unverified ASR; no playback; analytical windows, not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 12,
          "role": "hook",
          "summary": "Preview introduces the claimed strategic-time benefit.",
          "visual": "Title card followed by purple-headed account forecast graph.",
          "evidence_time": 5.5,
          "mechanics": "Finished-looking interface appears before the presenter explains the problem.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 3.04
          }
        },
        {
          "start": 12,
          "end": 64,
          "role": "context",
          "summary": "Presenter describes monthly forecasting and collaboration pressure.",
          "visual": "Seated presenter behind laptop against curtains and shelving.",
          "evidence_time": 40.5,
          "mechanics": "Repeated presenter samples with changing gestures replace screen detail.",
          "asr_anchor": {
            "word_index": 58,
            "source_time": 12.48
          }
        },
        {
          "start": 64,
          "end": 88,
          "role": "task",
          "summary": "Request combines actuals, baseline, and interactive forecasting.",
          "visual": "Chat conversation and resource sidebar; upper-left presenter tile.",
          "evidence_time": 70.5,
          "mechanics": "Screen share dominates while presenter remains a small inset.",
          "asr_anchor": {
            "word_index": 420,
            "source_time": 64.14
          }
        },
        {
          "start": 88,
          "end": 110,
          "role": "proof",
          "summary": "Aggregate output compares baseline and scenario trajectories.",
          "visual": "Purple-headed site with summary table and multi-line chart.",
          "evidence_time": 100.5,
          "mechanics": "Table-centered samples precede a larger chart-centered viewport.",
          "asr_anchor": {
            "word_index": 558,
            "source_time": 88.12
          }
        },
        {
          "start": 110,
          "end": 144.174,
          "role": "proof",
          "summary": "Account-level adjustment illustrates the claimed collaborative model.",
          "visual": "Account sidebar and diverging forecast lines; final pink dot.",
          "evidence_time": 130.5,
          "mechanics": "Later account sample shows a different green trajectory before white ending.",
          "asr_anchor": {
            "word_index": 724,
            "source_time": 110.02
          }
        }
      ],
      "editing_lesson": {
        "start": 110,
        "end": 138,
        "observation": "Same account chart shows different green forecast geometry across samples.",
        "inference": "Interpretation: stable framing makes the adjustment's local consequence comparable.",
        "adaptation": "Record your baseline, change one input, then show local and aggregate results.",
        "caution": "Samples do not show the claimed aggregate recalculation after adjustment.",
        "evidence_times": [
          115.5,
          125.5,
          130.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            2,
            3
          ]
        }
      ],
      "uncertainty": "No continuous drag review; end card beyond the sampled pink dot is unknown.",
      "media": {
        "sha256": "94be36de9b7e167295d824bdb5d97ae3a7a7836ab93b9f42056b001939473752",
        "duration": 144.174,
        "width": 854,
        "height": 480,
        "fps": "24/1",
        "retrieved_at": "2026-09-17T18:16:15.830878+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 464,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "94be36de9b7e167295d824bdb5d97ae3a7a7836ab93b9f42056b001939473752"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "6VwIXjFj6YQ",
      "rank": 41,
      "title": "You can just launch Sites | ChatGPT Work",
      "source_url": "https://www.youtube.com/watch?v=6VwIXjFj6YQ",
      "publication_date": "2026-08-13",
      "publication_timestamp": "2026-08-13T01:12:53-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A typed request sets up a launch-page task. Waiting and editing samples lead to a collage suggesting broader uses.",
      "family": "product-film",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "5-second stills; event-only ASR, no speech evidence; analytical windows, not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 5,
          "role": "hook",
          "summary": "Brand precedes the task.",
          "visual": "Small black knot logo on white.",
          "evidence_time": 0.5,
          "mechanics": "Wide empty margins isolate the centered mark.",
          "asr_anchor": null
        },
        {
          "start": 5,
          "end": 15,
          "role": "task",
          "summary": "A request specifies a shareable launch page.",
          "visual": "Enlarged composer with blue Sites reference.",
          "evidence_time": 10.5,
          "mechanics": "Fragment at 5.5 becomes longer, feature-tagged text at 10.5.",
          "asr_anchor": null
        },
        {
          "start": 15,
          "end": 23,
          "role": "proof",
          "summary": "The film supplies an editable-looking result.",
          "visual": "Waiting text, then purple board with formatting panel.",
          "evidence_time": 20.5,
          "mechanics": "Sparse status contrasts with a dense page-editing interface.",
          "asr_anchor": null
        },
        {
          "start": 23,
          "end": 28.594,
          "role": "close",
          "summary": "Other outputs broaden the pitch.",
          "visual": "Product name surrounded by colorful page fragments.",
          "evidence_time": 25.5,
          "mechanics": "Several peripheral examples share one white composition.",
          "asr_anchor": null
        }
      ],
      "editing_lesson": {
        "start": 5,
        "end": 23,
        "observation": "Request, waiting text, and purple editing interface appear in successive samples.",
        "inference": "Interpretation: the result answers the request visually.",
        "adaptation": "Record your own brief, real processing state, and resulting page editor.",
        "caution": "Samples cannot establish generation latency or editing responsiveness.",
        "evidence_times": [
          10.5,
          15.5,
          20.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "code": "output-montage",
          "beat_indices": [
            3
          ]
        }
      ],
      "uncertainty": "No speech anchor; processing and page-reveal choreography remain unseen.",
      "media": {
        "sha256": "f7bcac69138a6f32dfa2cff972ecbc596f996648b67e0c009fd8395237585a66",
        "duration": 28.594,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:16:14.434295+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 0,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "f7bcac69138a6f32dfa2cff972ecbc596f996648b67e0c009fd8395237585a66"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "-vvGAKtV_Ek",
      "rank": 42,
      "title": "You can just finish the work | ChatGPT Work",
      "source_url": "https://www.youtube.com/watch?v=-vvGAKtV_Ek",
      "publication_date": "2026-08-13",
      "publication_timestamp": "2026-08-13T01:09:08-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A casual request asks for a deck from mixed sources. Working status develops that promise, but sampled footage closes without showing the deck.",
      "family": "product-film",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "5-second stills; event-only ASR, no speech evidence; analytical windows, not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 5,
          "role": "hook",
          "summary": "A minimal brand opening.",
          "visual": "Black knot logo on white.",
          "evidence_time": 0.5,
          "mechanics": "Small central mark leaves extensive empty space.",
          "asr_anchor": null
        },
        {
          "start": 5,
          "end": 13,
          "role": "task",
          "summary": "Multiple sources feed a deck request.",
          "visual": "Composer references spreadsheet, PDF, and Slack material.",
          "evidence_time": 10.5,
          "mechanics": "Extreme text crop gives way to the whole composer.",
          "asr_anchor": null
        },
        {
          "start": 13,
          "end": 19,
          "role": "action",
          "summary": "The interface claims work is underway.",
          "visual": "Enlarged progress text and partially cropped response.",
          "evidence_time": 15.5,
          "mechanics": "Status replaces the source-filled input as focal content.",
          "asr_anchor": null
        },
        {
          "start": 19,
          "end": 26.854,
          "role": "close",
          "summary": "Slogan and branding replace execution detail.",
          "visual": "Bold slogan fragment, then knot logo.",
          "evidence_time": 25.5,
          "mechanics": "Both closing samples retain the sparse white field.",
          "asr_anchor": null
        }
      ],
      "editing_lesson": {
        "start": 5,
        "end": 19,
        "observation": "Composer framing expands to reveal sources before progress text occupies the screen.",
        "inference": "Interpretation: source context makes delegation legible.",
        "adaptation": "Show an original rough request, reveal its attachments, then actual execution status.",
        "caution": "This sequence alone cannot prove a usable deck exists.",
        "evidence_times": [
          5.5,
          10.5,
          15.5
        ]
      },
      "codes": [
        {
          "code": "context-before-detail",
          "beat_indices": [
            1,
            2
          ]
        }
      ],
      "uncertainty": "Speech absent in ASR; an intervening artifact could fall between samples.",
      "media": {
        "sha256": "dc322d160f06005f7292fca2ad2bbb37e1ec9d496e97d2de900a494bc489f3b0",
        "duration": 26.854,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:16:15.151888+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 0,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "dc322d160f06005f7292fca2ad2bbb37e1ec9d496e97d2de900a494bc489f3b0"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "OEkxKdhtQng",
      "rank": 43,
      "title": "How Model ML Uses GPT-5.6 Sol to Get Finance Work Done More Efficiently",
      "source_url": "https://www.youtube.com/watch?v=OEkxKdhtQng",
      "publication_date": "2026-08-10",
      "publication_timestamp": "2026-08-10T06:33:42-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "An interviewee contrasts disposable AI output with useful financial work. A memo-building example supports his pitch before he returns to human judgment.",
      "family": "testimonial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "5-second stills and unverified ASR only; windows are not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 24,
          "role": "hook",
          "summary": "The speaker stakes a quality claim.",
          "visual": "Seated interview, company card, dark website.",
          "evidence_time": 20.5,
          "mechanics": "Interview framing brackets introductory company material.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.12
          }
        },
        {
          "start": 24,
          "end": 40,
          "role": "task",
          "summary": "A brief becomes a proposed plan.",
          "visual": "Side-angle interview followed by white prompt panes.",
          "evidence_time": 35.5,
          "mechanics": "Human explanation yields to request and response samples.",
          "asr_anchor": {
            "word_index": 132,
            "source_time": 24.42
          }
        },
        {
          "start": 40,
          "end": 58,
          "role": "proof",
          "summary": "The film presents a chart-bearing deliverable.",
          "visual": "Checklist, valuation chart, and desktop file view.",
          "evidence_time": 55.5,
          "mechanics": "Chart appears standalone before appearing inside desktop chrome.",
          "asr_anchor": {
            "word_index": 266,
            "source_time": 40.08
          }
        },
        {
          "start": 58,
          "end": 83.854,
          "role": "qualification",
          "summary": "Judgment remains the user's responsibility.",
          "visual": "Alternating interview angles, then white wordmark card.",
          "evidence_time": 60.5,
          "mechanics": "Person replaces software for the concluding assessment.",
          "asr_anchor": {
            "word_index": 386,
            "source_time": 58.0
          }
        }
      ],
      "editing_lesson": {
        "start": 30,
        "end": 62,
        "observation": "Prompt pane precedes chart-bearing desktop, then a side-angle interview returns.",
        "inference": "Interpretation: the artifact grounds the judgment discussion.",
        "adaptation": "Capture a real brief, its editable deliverable, then explain remaining decisions.",
        "caution": "Traceability and quality remain testimonial claims, not audited results.",
        "evidence_times": [
          30.5,
          55.5,
          60.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            3
          ]
        }
      ],
      "uncertainty": "ASR varies the model name; chart accuracy is untested.",
      "media": {
        "sha256": "446b34b8d6af1330883b2455eb105356900f0fd6f5d83e91dcf3eed07e134291",
        "duration": 83.854,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:17.193637+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 258,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "446b34b8d6af1330883b2455eb105356900f0fd6f5d83e91dcf3eed07e134291"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "p_slDAvPjv0",
      "rank": 44,
      "title": "How to Schedule a Weekly Metrics Report With ChatGPT Work",
      "source_url": "https://www.youtube.com/watch?v=p_slDAvPjv0",
      "publication_date": "2026-08-06",
      "publication_timestamp": "2026-08-06T15:29:28-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A scientist describes recurring reporting effort. Setup, report review, and scheduling lead back to her claimed time savings.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "5-second stills and unverified ASR only; windows are not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 32,
          "role": "context",
          "summary": "Weekly reporting establishes the recurring burden.",
          "visual": "Presenter, laptop, small report overlay, blue mascot.",
          "evidence_time": 10.5,
          "mechanics": "Report appears beside her before setup instructions.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 3.58
          }
        },
        {
          "start": 32,
          "end": 50,
          "role": "action",
          "summary": "She explains opening Work and connecting data.",
          "visual": "Numbered instruction text over laptop-side footage.",
          "evidence_time": 40.5,
          "mechanics": "Large step labels precede the smaller source-request card.",
          "asr_anchor": {
            "word_index": 220,
            "source_time": 35.1
          }
        },
        {
          "start": 50,
          "end": 70,
          "role": "action",
          "summary": "Report review precedes recurring scheduling.",
          "visual": "KPI prompt, report deck, recurring-task prompt.",
          "evidence_time": 60.5,
          "mechanics": "Different overlays occupy the same laptop composition.",
          "asr_anchor": {
            "word_index": 300,
            "source_time": 51.28
          }
        },
        {
          "start": 70,
          "end": 93.854,
          "role": "close",
          "summary": "She restates the recurring-draft benefit.",
          "visual": "Presenter flanked by reports, then inset portrait.",
          "evidence_time": 75.5,
          "mechanics": "Output overlays disappear before the framed closing portrait.",
          "asr_anchor": {
            "word_index": 398,
            "source_time": 71.24
          }
        }
      ],
      "editing_lesson": {
        "start": 50,
        "end": 70,
        "observation": "Report prompt, slide-deck result, and scheduling prompt reuse the laptop background.",
        "inference": "Interpretation: checking one example precedes delegating repetition.",
        "adaptation": "Show your brief, inspect its generated report, then schedule that checked workflow.",
        "caution": "No future scheduled execution is sampled.",
        "evidence_times": [
          55.5,
          60.5,
          65.5
        ]
      },
      "codes": [
        {
          "code": "chapter-reset",
          "beat_indices": [
            1
          ]
        },
        {
          "code": "input-to-artifact",
          "beat_indices": [
            2
          ]
        }
      ],
      "uncertainty": "Demonstration data and reported savings are not independently validated.",
      "media": {
        "sha256": "a186c50ff6c81e4438749c46daf22f7f5cee2fb95abe6ba01470e2d7817faa31",
        "duration": 93.854,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:16.771410+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 248,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "a186c50ff6c81e4438749c46daf22f7f5cee2fb95abe6ba01470e2d7817faa31"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "TryfaGZwvIE",
      "rank": 45,
      "title": "How to Turn a Forecast Spreadsheet Into an Interactive Planning Tool With ChatGPT Work",
      "source_url": "https://www.youtube.com/watch?v=TryfaGZwvIE",
      "publication_date": "2026-08-06",
      "publication_timestamp": "2026-08-06T15:29:01-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Repeated forecast revisions motivate an interactive review. Source context and scenario instructions culminate in a result checked before spreadsheet approval.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "5-second stills, approval-frame inspection, and unverified ASR; windows are not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 32,
          "role": "context",
          "summary": "The presenter frames repeated finance-review work.",
          "visual": "Forecast preview beside presenter; colorful mascot.",
          "evidence_time": 10.5,
          "mechanics": "Output preview precedes sustained talking-head context.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 3.38
          }
        },
        {
          "start": 32,
          "end": 58,
          "role": "task",
          "summary": "He supplies spreadsheet context and decision requirements.",
          "visual": "Numbered opening step, then successive prompt cards.",
          "evidence_time": 55.5,
          "mechanics": "Source reference precedes audience-and-decision instructions.",
          "asr_anchor": {
            "word_index": 168,
            "source_time": 32.02
          }
        },
        {
          "start": 58,
          "end": 78,
          "role": "action",
          "summary": "Scenario instructions preserve the source spreadsheet.",
          "visual": "Multi-item site request followed by presenter.",
          "evidence_time": 65.5,
          "mechanics": "Longer request sample exposes constraints absent in the earlier crop.",
          "asr_anchor": {
            "word_index": 314,
            "source_time": 58.06
          }
        },
        {
          "start": 78,
          "end": 96.494,
          "role": "qualification",
          "summary": "He describes checking results before approving changes.",
          "visual": "Forecast chart, approval-status crop, presenter, logo.",
          "evidence_time": 85.5,
          "mechanics": "Broad result view precedes a focused approval area.",
          "asr_anchor": {
            "word_index": 412,
            "source_time": 78.42
          }
        }
      ],
      "editing_lesson": {
        "start": 60,
        "end": 88,
        "observation": "Constraint-bearing request precedes full forecast and cropped approval status.",
        "inference": "Interpretation: approval becomes part of the payoff.",
        "adaptation": "Record a protected-source request, inspect its scenario result, then show explicit approval.",
        "caution": "The still shows approval status, not the click or source write.",
        "evidence_times": [
          65.5,
          80.5,
          85.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            2,
            3
          ]
        },
        {
          "code": "permission-step",
          "beat_indices": [
            3
          ]
        }
      ],
      "uncertainty": "Forecast arithmetic and actual spreadsheet mutation remain unverified.",
      "media": {
        "sha256": "709f7303de6b3c634650776c32e6eb2522c9a3a519754bd23ea93ccbc91006d9",
        "duration": 96.494,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:17.742811+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 244,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "709f7303de6b3c634650776c32e6eb2522c9a3a519754bd23ea93ccbc91006d9"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "UaeWJK_vv-Y",
      "rank": 46,
      "title": "Introducing Agent Plugins",
      "source_url": "https://www.youtube.com/watch?v=UaeWJK_vv-Y",
      "publication_date": "2026-08-06",
      "publication_timestamp": "2026-08-06T09:13:08-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A presenter identifies incompatible agent-extension packaging. Manifest examples develop a shared-format proposal while the ending explicitly limits its scope.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "5-second stills and unverified ASR only; windows are not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 20,
          "role": "context",
          "summary": "The presenter describes fragmented packaging.",
          "visual": "Laptop close-up and two interview framings.",
          "evidence_time": 15.5,
          "mechanics": "Object introduction gives way to the seated explainer.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 3.92
          }
        },
        {
          "start": 20,
          "end": 38,
          "role": "context",
          "summary": "The film introduces a collaborative specification.",
          "visual": "Dark title, supporter logos, documentation page.",
          "evidence_time": 35.5,
          "mechanics": "Sparse cards precede dense technical documentation.",
          "asr_anchor": {
            "word_index": 88,
            "source_time": 20.16
          }
        },
        {
          "start": 38,
          "end": 62,
          "role": "action",
          "summary": "Folder and manifest examples explain packaging.",
          "visual": "Code editor with presenter inset; full presenter returns.",
          "evidence_time": 50.5,
          "mechanics": "Folder contents differ while the editor composition persists.",
          "asr_anchor": {
            "word_index": 172,
            "source_time": 38.06
          }
        },
        {
          "start": 62,
          "end": 95.114,
          "role": "qualification",
          "summary": "He separates packaging from permissions and runtimes.",
          "visual": "Expanded manifest, presenter, white end card.",
          "evidence_time": 75.5,
          "mechanics": "Denser code precedes the presenter-led scope qualification.",
          "asr_anchor": {
            "word_index": 368,
            "source_time": 76.12
          }
        }
      ],
      "editing_lesson": {
        "start": 35,
        "end": 57,
        "observation": "Documentation becomes an editor with presenter inset, then full presenter.",
        "inference": "Interpretation: implementation detail retains an identifiable guide.",
        "adaptation": "Introduce your format, show a minimal working folder, then explain its boundary.",
        "caution": "These examples do not demonstrate compatibility across products.",
        "evidence_times": [
          35.5,
          40.5,
          55.5
        ]
      },
      "codes": [
        {
          "code": "context-before-detail",
          "beat_indices": [
            0,
            1,
            2
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            2,
            3
          ]
        }
      ],
      "uncertainty": "No installation or runtime execution is established by the samples.",
      "media": {
        "sha256": "89eb898a0357dffdbdec3826785bd998ac0bf9323a449a03e73a17449023cdef",
        "duration": 95.114,
        "width": 854,
        "height": 480,
        "fps": "30000/1001",
        "retrieved_at": "2026-09-17T18:16:18.000037+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 221,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "89eb898a0357dffdbdec3826785bd998ac0bf9323a449a03e73a17449023cdef"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "SJJBPwlPpww",
      "rank": 47,
      "title": "How AI Helps Solve Medical Mysteries at Boston Children’s Hospital | OpenAI Forum",
      "source_url": "https://www.youtube.com/watch?v=SJJBPwlPpww",
      "publication_date": "2026-08-04",
      "publication_timestamp": "2026-08-04T13:29:24-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A participant's diagnostic search gives the research discussion personal stakes. Expert explanations develop reported findings, then audience questions expose unresolved validation and access issues.",
      "family": "interview",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "60-second stills and complete unverified ASR only; analytical windows, not continuous playback."
      },
      "beats": [
        {
          "start": 0,
          "end": 104,
          "role": "context",
          "summary": "Moderator introduces research and reported diagnostic leads.",
          "visual": "Event card, then solo moderator on blue-pink layout.",
          "evidence_time": 60.5,
          "mechanics": "Portrait roster gives way to one enlarged webcam.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 9.68
          }
        },
        {
          "start": 104,
          "end": 335,
          "role": "discussion",
          "summary": "Participant recounts delays and diagnosis's personal consequences.",
          "visual": "Moderator and participant in paired webcam boxes.",
          "evidence_time": 240.5,
          "mechanics": "Stable two-person arrangement keeps interviewer and subject visible.",
          "asr_anchor": {
            "word_index": 520,
            "source_time": 105.72
          }
        },
        {
          "start": 335,
          "end": 967,
          "role": "context",
          "summary": "Experts describe analysis burdens and family experiences.",
          "visual": "Four labeled webcams in a fixed grid.",
          "evidence_time": 540.5,
          "mechanics": "Panel replaces the pair; positions persist across samples.",
          "asr_anchor": {
            "word_index": 2601,
            "source_time": 482.02
          }
        },
        {
          "start": 967,
          "end": 1405,
          "role": "discussion",
          "summary": "Researchers explain filtered inputs, hypotheses, and expert checking.",
          "visual": "Same four-person grid; no workflow diagram sampled.",
          "evidence_time": 1200.5,
          "mechanics": "Explanations change without a sampled layout reset.",
          "asr_anchor": {
            "word_index": 5521,
            "source_time": 997.167
          }
        },
        {
          "start": 1405,
          "end": 1765,
          "role": "qualification",
          "summary": "Reported diagnoses differ from exploratory leads; access remains unfinished.",
          "visual": "Panelists remain visible together.",
          "evidence_time": 1500.5,
          "mechanics": "Shared grid persists through results and deployment discussion.",
          "asr_anchor": {
            "word_index": 7951,
            "source_time": 1431.328
          }
        },
        {
          "start": 1765,
          "end": 1995,
          "role": "discussion",
          "summary": "Speakers discuss reanalysis hopes, unsolved cases, and funding provenance.",
          "visual": "Four webcams against unchanged event graphics.",
          "evidence_time": 1860.5,
          "mechanics": "No sampled visual separation distinguishes forecasts from findings.",
          "asr_anchor": {
            "word_index": 9843,
            "source_time": 1765.5
          }
        },
        {
          "start": 1995,
          "end": 2295,
          "role": "qualification",
          "summary": "Audience questions probe sequencing, missed patterns, and reasoning versus curation.",
          "visual": "Panel grid; no question cards sampled.",
          "evidence_time": 2220.5,
          "mechanics": "Question changes are carried by ASR, not new graphics.",
          "asr_anchor": {
            "word_index": 11175,
            "source_time": 2004.208
          }
        },
        {
          "start": 2295,
          "end": 2371.3,
          "role": "close",
          "summary": "Moderator thanks participants and promotes continuing research.",
          "visual": "Solo moderator enlarged within the event frame.",
          "evidence_time": 2340.5,
          "mechanics": "Single webcam replaces the preceding panel sample.",
          "asr_anchor": {
            "word_index": 13063,
            "source_time": 2323.124
          }
        }
      ],
      "editing_lesson": {
        "start": 60,
        "end": 421,
        "observation": "Solo moderator becomes a participant pair, then a four-person expert grid.",
        "inference": "Interpretation: personal consequences precede technical breadth.",
        "adaptation": "Introduce a real user, explore their difficulty, then widen to responsible specialists.",
        "caution": "Minute sampling cannot establish handoff timing; panel testimony does not validate diagnoses.",
        "evidence_times": [
          60.5,
          180.5,
          360.5
        ]
      },
      "codes": [
        {
          "code": "human-stakes",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "code": "question-led",
          "beat_indices": [
            1,
            3,
            6
          ]
        }
      ],
      "uncertainty": "Minute gaps conceal edits; medical findings remain speaker-attributed, including unresolved leads.",
      "media": {
        "sha256": "858aab20c499872aeb07b75965a521c966a4c1c970b0127c0bba3880fb73300b",
        "duration": 2371.3,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:16:26.416550+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 6644,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "858aab20c499872aeb07b75965a521c966a4c1c970b0127c0bba3880fb73300b"
      },
      "sampling_interval": 60,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "r64krUavXJU",
      "rank": 48,
      "title": "Meet Birding Pal",
      "source_url": "https://www.youtube.com/watch?v=r64krUavXJU",
      "publication_date": "2026-08-04",
      "publication_timestamp": "2026-08-04T10:43:45-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A plush bird is presented as an outdoor identification companion. Parts and laptop samples precede field exchanges, suggesting a build-to-use arc.",
      "family": "product-film",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "5-second stills and unverified ASR only; windows are not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 10,
          "role": "hook",
          "summary": "A toy offers a bird identification.",
          "visual": "Handheld plush bird above an outdoor field guide.",
          "evidence_time": 5.5,
          "mechanics": "Toy and guide share the nature setting.",
          "asr_anchor": {
            "word_index": 2,
            "source_time": 2.22
          }
        },
        {
          "start": 10,
          "end": 27,
          "role": "action",
          "summary": "A construction montage suggests the toy's implementation.",
          "visual": "Parts layout, coding laptop, toy with circuitry overlay.",
          "evidence_time": 25.5,
          "mechanics": "Separated components precede the assembled-looking object.",
          "asr_anchor": null
        },
        {
          "start": 27,
          "end": 45,
          "role": "proof",
          "summary": "Field conversation proposes identities and reports logged birds.",
          "visual": "Binocular user, clipped-on toy, outdoor handheld close-up.",
          "evidence_time": 35.5,
          "mechanics": "Wide human context alternates with companion-focused samples.",
          "asr_anchor": {
            "word_index": 22,
            "source_time": 27.28
          }
        },
        {
          "start": 45,
          "end": 55.114,
          "role": "close",
          "summary": "The demonstration closes with project information.",
          "visual": "Toy in hands, then white destination cards.",
          "evidence_time": 50.5,
          "mechanics": "Outdoor texture gives way to sparse typography.",
          "asr_anchor": null
        }
      ],
      "editing_lesson": {
        "start": 15,
        "end": 36,
        "observation": "Parts mat precedes circuitry-overlay toy and outdoor wearable use.",
        "inference": "Interpretation: construction earns the field demonstration.",
        "adaptation": "Show your components, completed prototype, then a recorded user interaction.",
        "caution": "The montage cannot prove wiring correctness or identification accuracy.",
        "evidence_times": [
          15.5,
          25.5,
          35.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            1,
            2
          ]
        }
      ],
      "uncertainty": "Build and closing windows lack speech; call recognition is unverified.",
      "media": {
        "sha256": "5cf4314e91d697b773321e8732c438ad8abad4c88ced84903d9e80f75228fa22",
        "duration": 55.114,
        "width": 854,
        "height": 480,
        "fps": "30000/1001",
        "retrieved_at": "2026-09-17T18:16:19.122588+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 52,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "5cf4314e91d697b773321e8732c438ad8abad4c88ced84903d9e80f75228fa22"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "_Gd9yzAc-WI",
      "rank": 49,
      "title": "Using Voice in ChatGPT Work",
      "source_url": "https://www.youtube.com/watch?v=_Gd9yzAc-WI",
      "publication_date": "2026-07-29",
      "publication_timestamp": "2026-07-29T15:37:43-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "A presenter starts with unfinished tasks and a spoken brainstorming request. Cross-app context develops parallel deck and travel work, resolving with sharing and a purchase boundary.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "5-second stills and unverified ASR only; windows are not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 44,
          "role": "task",
          "summary": "A to-do list prompts event-pitch brainstorming.",
          "visual": "Notes list, presenter, laptop with voice orb.",
          "evidence_time": 10.5,
          "mechanics": "List establishes tasks before presenter-led ideation.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.44
          }
        },
        {
          "start": 44,
          "end": 74,
          "role": "context",
          "summary": "He distinguishes ongoing conversation from dictation.",
          "visual": "Close and wide presenter views at desk.",
          "evidence_time": 60.5,
          "mechanics": "Explanation uses human framing rather than a new artifact.",
          "asr_anchor": {
            "word_index": 278,
            "source_time": 45.36
          }
        },
        {
          "start": 74,
          "end": 128,
          "role": "action",
          "summary": "Slack and calendar supply travel context.",
          "visual": "Slack thread, then several calendar views with orb.",
          "evidence_time": 120.5,
          "mechanics": "Persistent orb accompanies different app surfaces.",
          "asr_anchor": {
            "word_index": 504,
            "source_time": 78.36
          }
        },
        {
          "start": 128,
          "end": 145,
          "role": "qualification",
          "summary": "Travel preparation explicitly excludes purchasing.",
          "visual": "Presenter close-up and Slack-over-laptop view.",
          "evidence_time": 135.5,
          "mechanics": "Purchase boundary is spoken; no approval dialog sampled here.",
          "asr_anchor": {
            "word_index": 800,
            "source_time": 128.0
          }
        },
        {
          "start": 145,
          "end": 177,
          "role": "proof",
          "summary": "Deck sharing and staged travel answer earlier tasks.",
          "visual": "Pitch slide, Slack message, flight and hotel summaries.",
          "evidence_time": 175.5,
          "mechanics": "Distinct deliverables follow the earlier task branches.",
          "asr_anchor": {
            "word_index": 930,
            "source_time": 145.32
          }
        },
        {
          "start": 177,
          "end": 193.414,
          "role": "close",
          "summary": "Presenter resumes control and closes the walkthrough.",
          "visual": "Wide presenter, close-up, then white wordmark.",
          "evidence_time": 185.5,
          "mechanics": "Software views yield to the person and branding.",
          "asr_anchor": {
            "word_index": 1128,
            "source_time": 177.28
          }
        }
      ],
      "editing_lesson": {
        "start": 150,
        "end": 177,
        "observation": "Pitch slide, sent-message view, and travel-review screens supply separate task payoffs.",
        "inference": "Interpretation: returning to both branches makes multitasking legible.",
        "adaptation": "Seed two original requests, show each result, and retain its authorization boundary.",
        "caution": "Staged booking screens do not establish background autonomy or purchase completion.",
        "evidence_times": [
          150.5,
          160.5,
          175.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            0,
            4
          ]
        },
        {
          "code": "permission-step",
          "beat_indices": [
            3,
            4
          ]
        }
      ],
      "uncertainty": "ASR's coding-product wording is uncertain; depicted calendar dates may not align with narration.",
      "media": {
        "sha256": "ca73dbcd553720c2688733dc185f3d4fbef7b619d1e31ab339fd65af513a153a",
        "duration": 193.414,
        "width": 854,
        "height": 480,
        "fps": "30000/1001",
        "retrieved_at": "2026-09-17T18:16:19.811316+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 610,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "ca73dbcd553720c2688733dc185f3d4fbef7b619d1e31ab339fd65af513a153a"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "MLehRytu9Zo",
      "rank": 50,
      "title": "We're giving 100,000 academic researchers free access to our frontier models",
      "source_url": "https://www.youtube.com/watch?v=MLehRytu9Zo",
      "publication_date": "2026-07-29",
      "publication_timestamp": "2026-07-29T10:23:02-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Four researchers discuss what AI changes about scientific inquiry. Personal examples broaden into hopes for distributed access, leaving future discoveries unspecified.",
      "family": "interview",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "5-second stills and unverified ASR only; windows are not cuts."
      },
      "beats": [
        {
          "start": 0,
          "end": 34,
          "role": "context",
          "summary": "Personal curiosity introduces the researchers.",
          "visual": "Overhead meeting room, arrivals, seated conversation.",
          "evidence_time": 15.5,
          "mechanics": "Room geography precedes individual and paired portraits.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.32
          }
        },
        {
          "start": 34,
          "end": 84,
          "role": "discussion",
          "summary": "Speakers recount analogy and data-analysis examples.",
          "visual": "Women speaking, listeners, hands beside notebooks.",
          "evidence_time": 65.5,
          "mechanics": "Close-ups and table details vary the shared setting.",
          "asr_anchor": {
            "word_index": 218,
            "source_time": 37.06
          }
        },
        {
          "start": 84,
          "end": 138,
          "role": "discussion",
          "summary": "Participants speculate about conjectures and faster discovery.",
          "visual": "Long-haired participant, then beige-shirted speaker and group.",
          "evidence_time": 115.5,
          "mechanics": "Single-speaker framing expands to include listeners.",
          "asr_anchor": {
            "word_index": 522,
            "source_time": 85.52
          }
        },
        {
          "start": 138,
          "end": 185,
          "role": "discussion",
          "summary": "Time for research motivates wider access.",
          "visual": "Individual reactions and four-person table views.",
          "evidence_time": 155.5,
          "mechanics": "Group compositions recur alongside speaker close-ups.",
          "asr_anchor": {
            "word_index": 954,
            "source_time": 138.42
          }
        },
        {
          "start": 185,
          "end": 210.194,
          "role": "close",
          "summary": "Curiosity closes an open-ended outlook.",
          "visual": "Blue-shirted speaker, empty overhead table, logo.",
          "evidence_time": 205.5,
          "mechanics": "Occupied conversation gives way to the emptied establishing space.",
          "asr_anchor": {
            "word_index": 1284,
            "source_time": 185.0
          }
        }
      ],
      "editing_lesson": {
        "start": 0,
        "end": 210.194,
        "observation": "Early overhead room and gathered group contrast with an empty table near closing.",
        "inference": "Interpretation: the shared place bookends the exchange.",
        "adaptation": "Film your meeting space, the actual discussion, and that space after participants leave.",
        "caution": "Spatial closure does not substantiate the speakers' discovery forecasts.",
        "evidence_times": [
          0.5,
          15.5,
          205.5
        ]
      },
      "codes": [
        {
          "code": "human-stakes",
          "beat_indices": [
            0,
            3,
            4
          ]
        }
      ],
      "uncertainty": "No research artifacts or access-allocation details are established by these samples.",
      "media": {
        "sha256": "12bfb81016bd4ec4c2a18597c842fb90f3c039f5880d63b8c6e00e9ff0c347ba",
        "duration": 210.194,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:19.581549+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 697,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "12bfb81016bd4ec4c2a18597c842fb90f3c039f5880d63b8c6e00e9ff0c347ba"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "WeP9VUf1OoE",
      "rank": 51,
      "title": "Introducing gpt-transcribe and gpt-live-transcribe",
      "source_url": "https://www.youtube.com/watch?v=WeP9VUf1OoE",
      "publication_date": "2026-07-28",
      "publication_timestamp": "2026-07-28T14:49:18-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "The presenter distinguishes live transcription from completed-file processing. Vocabulary, background-distraction and language-switch examples precede a batch-processing result, resolving the distinction through contrasting workflows.",
      "family": "product-film",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second still samples and unverified ASR; no playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 30,
          "role": "context",
          "summary": "Presenter introduces two transcription workflows.",
          "visual": "Presenter beside laptop; enlarged transcript screen.",
          "evidence_time": 15.5,
          "mechanics": "Wide desk framing contrasts with screen-only samples.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.22
          }
        },
        {
          "start": 30,
          "end": 75,
          "role": "proof",
          "summary": "Film demonstrates vocabulary and distraction handling.",
          "visual": "Transcript closeups; presenter holds a small instrument.",
          "evidence_time": 60.5,
          "mechanics": "Prop-bearing presenter differs from subsequent text-only samples.",
          "asr_anchor": {
            "word_index": 142,
            "source_time": 32.119
          }
        },
        {
          "start": 75,
          "end": 96,
          "role": "proof",
          "summary": "ASR records a bilingual demonstration.",
          "visual": "Spanish text appears on the laptop.",
          "evidence_time": 90.5,
          "mechanics": "Text closeups sit between repeated presenter framings.",
          "asr_anchor": {
            "word_index": 414,
            "source_time": 81.32
          }
        },
        {
          "start": 96,
          "end": 114,
          "role": "proof",
          "summary": "Film presents completed-file processing and its result.",
          "visual": "Timer interface, then completion display.",
          "evidence_time": 110.5,
          "mechanics": "Two sampled timer states bracket a presenter image.",
          "asr_anchor": {
            "word_index": 506,
            "source_time": 98.58
          }
        },
        {
          "start": 114,
          "end": 131.314,
          "role": "close",
          "summary": "Presenter assigns each workflow suitable uses.",
          "visual": "Presenter and laptop remain at desk.",
          "evidence_time": 125.5,
          "mechanics": "Closing samples retain the original wide composition.",
          "asr_anchor": {
            "word_index": 634,
            "source_time": 119.26
          }
        }
      ],
      "editing_lesson": {
        "start": 96,
        "end": 120,
        "observation": "Timer, completion display and presenter occupy successive sampled states.",
        "inference": "Interpretation: the result supports the closing workflow distinction.",
        "adaptation": "Record an original upload, result and presenter explanation.",
        "caution": "The timer jump cannot establish actual processing speed.",
        "evidence_times": [
          100.5,
          110.5,
          115.5
        ]
      },
      "codes": [
        {
          "code": "presenter-return",
          "beat_indices": [
            0,
            2,
            3,
            4
          ]
        },
        {
          "code": "proof-before-close",
          "beat_indices": [
            3,
            4
          ]
        }
      ],
      "uncertainty": "Noise rejection is asserted, not audibly checked; sampled transcript errors remain visible.",
      "media": {
        "sha256": "a7b2610eb4f4c6eca30ec61640107842518f07bd099303a16e2189a70f677c22",
        "duration": 131.314,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:20.830149+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 343,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "a7b2610eb4f4c6eca30ec61640107842518f07bd099303a16e2189a70f677c22"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "g-g6E-qtCjA",
      "rank": 52,
      "title": "ChatGPT Work for Sales: Revenue Intelligence for Sales Leaders",
      "source_url": "https://www.youtube.com/watch?v=g-g6E-qtCjA",
      "publication_date": "2026-07-28",
      "publication_timestamp": "2026-07-28T07:25:26-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "The film asks how scattered sales signals become leadership decisions. A report develops into conversational investigation and a Slack draft, while continuous improvement remains a closing claim.",
      "family": "product-film",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second still samples and unverified ASR; no playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 30,
          "role": "context",
          "summary": "Film introduces a demonstration revenue report.",
          "visual": "Title cards, chat, desktop report with phone overlay.",
          "evidence_time": 20.5,
          "mechanics": "Device views replace the opening text-only proposition.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.18
          }
        },
        {
          "start": 30,
          "end": 83,
          "role": "proof",
          "summary": "Report connects trends, objections and recommended responses.",
          "visual": "Charts, source drawer and dark recommendation cards.",
          "evidence_time": 50.5,
          "mechanics": "Dense report samples include selectively isolated details.",
          "asr_anchor": {
            "word_index": 180,
            "source_time": 31.02
          }
        },
        {
          "start": 83,
          "end": 123,
          "role": "action",
          "summary": "Narrator investigates objections and requests leadership proposals.",
          "visual": "Chat answers include tables and bulleted recommendations.",
          "evidence_time": 105.5,
          "mechanics": "Conversational text replaces the formatted report surface.",
          "asr_anchor": {
            "word_index": 506,
            "source_time": 84.14
          }
        },
        {
          "start": 123,
          "end": 148.894,
          "role": "close",
          "summary": "Slack handoff precedes the promised learning cycle.",
          "visual": "Dark Slack draft, then white closing cards.",
          "evidence_time": 130.5,
          "mechanics": "Application density gives way to sparse typography.",
          "asr_anchor": {
            "word_index": 746,
            "source_time": 127.14
          }
        }
      ],
      "editing_lesson": {
        "start": 30,
        "end": 83,
        "observation": "Bars, a supporting-records drawer and a recommendation card appear in separate samples.",
        "inference": "Interpretation: investigation precedes prescription.",
        "adaptation": "Show an original finding, its source and an owner-ready response.",
        "caution": "The displayed sources do not independently validate recommendations.",
        "evidence_times": [
          40.5,
          50.5,
          75.5
        ]
      },
      "codes": [
        {
          "code": "context-before-detail",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "code": "input-to-artifact",
          "beat_indices": [
            2,
            3
          ]
        }
      ],
      "uncertainty": "No subsequent business results are sampled; the feedback cycle remains asserted.",
      "media": {
        "sha256": "9c889760d4211203b32a347a138b708618b89ed1a1e15773264d65ddbebe9c8d",
        "duration": 148.894,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:16:22.451115+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 431,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "9c889760d4211203b32a347a138b708618b89ed1a1e15773264d65ddbebe9c8d"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "5DbZAMGZ5oM",
      "rank": 53,
      "title": "ChatGPT Work for Sales: Account Research and Customer Outreach",
      "source_url": "https://www.youtube.com/watch?v=5DbZAMGZ5oM",
      "publication_date": "2026-07-28",
      "publication_timestamp": "2026-07-28T07:23:31-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "An account signal creates the problem of carrying customer context through selling. Source material, a brief, outreach and meeting assets lead to a narrated follow-up workflow rather than demonstrated commercial success.",
      "family": "product-film",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second still samples and unverified ASR; no playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 23,
          "role": "context",
          "summary": "An earnings signal initiates account research.",
          "visual": "Opening cards, Slack message and earnings PDF.",
          "evidence_time": 15.5,
          "mechanics": "Source document follows the workflow proposition.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.12
          }
        },
        {
          "start": 23,
          "end": 60,
          "role": "action",
          "summary": "Brief organizes evidence, stakeholders and outreach.",
          "visual": "Account site contains profiles and email copy.",
          "evidence_time": 45.5,
          "mechanics": "Overview, excerpt drawer and profile cards differ across samples.",
          "asr_anchor": {
            "word_index": 172,
            "source_time": 25.88
          }
        },
        {
          "start": 60,
          "end": 82,
          "role": "proof",
          "summary": "Narrated booking triggers customer-specific meeting preparation.",
          "visual": "Slack notification precedes presentation editor slides.",
          "evidence_time": 70.5,
          "mechanics": "Message context gives way to designed meeting artifacts.",
          "asr_anchor": {
            "word_index": 396,
            "source_time": 62.44
          }
        },
        {
          "start": 82,
          "end": 111.574,
          "role": "close",
          "summary": "Post-call follow-up closes the claimed workflow.",
          "visual": "Slack thread, closing cards and logo.",
          "evidence_time": 90.5,
          "mechanics": "Expanded thread contrasts with the sparse ending.",
          "asr_anchor": {
            "word_index": 538,
            "source_time": 84.62
          }
        }
      ],
      "editing_lesson": {
        "start": 10,
        "end": 45,
        "observation": "An earnings PDF precedes an account site and source-excerpt drawer.",
        "inference": "Interpretation: visible provenance makes the recommendation easier to inspect.",
        "adaptation": "Record an original document, derived brief and linked supporting passage.",
        "caution": "Source visibility does not verify the brief's interpretation.",
        "evidence_times": [
          15.5,
          25.5,
          40.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            0,
            1,
            2
          ]
        }
      ],
      "uncertainty": "ASR mentions Gmail; sampled outreach stays in the brief and Slack.",
      "media": {
        "sha256": "44b914a8d665a2b813a50650ed1171608e6df2cf3f1808969063dadc9f75c325",
        "duration": 111.574,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:16:21.909486+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 338,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "44b914a8d665a2b813a50650ed1171608e6df2cf3f1808969063dadc9f75c325"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "hjXi7b01s2c",
      "rank": 54,
      "title": "ChatGPT Work for Sales: Account Planning and Meeting Prep",
      "source_url": "https://www.youtube.com/watch?v=hjXi7b01s2c",
      "publication_date": "2026-07-28",
      "publication_timestamp": "2026-07-28T07:20:18-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "The narrator frames inconsistent sales execution as a preparation problem. A daily-planning request develops into workshop context, reusable assets and a team handoff, supporting the film's consistency promise.",
      "family": "product-film",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second still samples and unverified ASR; no playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 32,
          "role": "task",
          "summary": "Seller requests a prioritized daily plan.",
          "visual": "Opening cards, empty composer and returned schedule.",
          "evidence_time": 30.5,
          "mechanics": "Sparse input expands into a structured response.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.22
          }
        },
        {
          "start": 32,
          "end": 62,
          "role": "context",
          "summary": "Upcoming workshop motivates deeper account preparation.",
          "visual": "Agenda beside account dashboard and stakeholder diagram.",
          "evidence_time": 45.5,
          "mechanics": "Overlapping windows connect schedule and account detail.",
          "asr_anchor": {
            "word_index": 174,
            "source_time": 34.34
          }
        },
        {
          "start": 62,
          "end": 84,
          "role": "proof",
          "summary": "Film presents generated workshop materials.",
          "visual": "Deck, account site and embedded video player.",
          "evidence_time": 80.5,
          "mechanics": "Dark visual motifs recur across different artifact surfaces.",
          "asr_anchor": {
            "word_index": 322,
            "source_time": 63.14
          }
        },
        {
          "start": 84,
          "end": 105.314,
          "role": "close",
          "summary": "Materials reach the account team's workspace.",
          "visual": "Chat, Slack channel and closing cards.",
          "evidence_time": 90.5,
          "mechanics": "Team-channel sample precedes the concluding promise.",
          "asr_anchor": {
            "word_index": 460,
            "source_time": 87.02
          }
        }
      ],
      "editing_lesson": {
        "start": 62,
        "end": 84,
        "observation": "Related dark graphics recur in a deck, site and video player.",
        "inference": "Interpretation: repeated design links separate outputs to one workshop.",
        "adaptation": "Carry original customer context through slides, a hub and a short demo.",
        "caution": "Player stills cannot establish the claimed animation quality.",
        "evidence_times": [
          65.5,
          75.5,
          80.5
        ]
      },
      "codes": [
        {
          "code": "input-to-artifact",
          "beat_indices": [
            0,
            1,
            2
          ]
        },
        {
          "code": "output-montage",
          "beat_indices": [
            2
          ]
        }
      ],
      "uncertainty": "Asset generation and customer response are not continuously observed.",
      "media": {
        "sha256": "51b0aeb16c22bf7286e725de924c39fb0a6cfb33b8d12733f09e9411e66132bc",
        "duration": 105.314,
        "width": 852,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:16:22.325368+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 276,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "51b0aeb16c22bf7286e725de924c39fb0a6cfb33b8d12733f09e9411e66132bc"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "q9TPE7pKkIo",
      "rank": 55,
      "title": "ChatGPT Work for Sales: Pipeline and Forecast Intelligence",
      "source_url": "https://www.youtube.com/watch?v=q9TPE7pKkIo",
      "publication_date": "2026-07-28",
      "publication_timestamp": "2026-07-28T07:16:53-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Leadership must choose which business signals deserve coordinated action. An executive brief develops into account investigation and a reviewed team message, ending with a promise of continued monitoring.",
      "family": "product-film",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second still samples and unverified ASR; no playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 30,
          "role": "context",
          "summary": "Executive brief prioritizes areas needing attention.",
          "visual": "Title cards, chat brief and isolated priority row.",
          "evidence_time": 15.5,
          "mechanics": "Dimmed surroundings distinguish one row from dense text.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.18
          }
        },
        {
          "start": 30,
          "end": 76,
          "role": "proof",
          "summary": "Dashboard develops a proposed sales initiative.",
          "visual": "Priority list, funnel chart and account drawer.",
          "evidence_time": 65.5,
          "mechanics": "Broad dashboard and enlarged account details alternate across samples.",
          "asr_anchor": {
            "word_index": 190,
            "source_time": 33.0
          }
        },
        {
          "start": 76,
          "end": 110,
          "role": "action",
          "summary": "Team preparation includes reviewing a drafted message.",
          "visual": "Slack membership modal, composer and attached plan.",
          "evidence_time": 95.5,
          "mechanics": "Setup, draft and plan occupy distinct sampled states.",
          "asr_anchor": {
            "word_index": 442,
            "source_time": 79.16
          }
        },
        {
          "start": 110,
          "end": 129.934,
          "role": "close",
          "summary": "Narration returns to ongoing monitoring.",
          "visual": "Channel messages precede white concluding cards.",
          "evidence_time": 120.5,
          "mechanics": "Detailed workspace yields to a short coordination promise.",
          "asr_anchor": {
            "word_index": 676,
            "source_time": 118.54
          }
        }
      ],
      "editing_lesson": {
        "start": 76,
        "end": 116,
        "observation": "Membership setup, a send-ready composer and later channel content remain distinguishable.",
        "inference": "Interpretation: review separates recommendation from execution.",
        "adaptation": "Show an original team setup, reviewable draft and confirmed post.",
        "caution": "Still samples do not capture the exact send action.",
        "evidence_times": [
          85.5,
          105.5,
          110.5
        ]
      },
      "codes": [
        {
          "code": "context-before-detail",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "code": "permission-step",
          "beat_indices": [
            2
          ]
        }
      ],
      "uncertainty": "Forecast probabilities and source reliability are film claims, not independently checked results.",
      "media": {
        "sha256": "1711351af922622355ae0955f818953579ebe7eb7c823cbe6578210b74282f22",
        "duration": 129.934,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:16:22.554395+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 364,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "1711351af922622355ae0955f818953579ebe7eb7c823cbe6578210b74282f22"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "oXu6JGxW_YM",
      "rank": 56,
      "title": "Navigating AI Transformation in the Legal Industry | Jonathan Williams, Head of France, Legora",
      "source_url": "https://www.youtube.com/watch?v=oXu6JGxW_YM",
      "publication_date": "2026-07-27",
      "publication_timestamp": "2026-07-27T05:37:15-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "The speaker's enthusiasm for model releases encounters the difficulty of enterprise change. Partnership testimony gives way to organizational constraints, leaving the future uncertain and adaptation as his proposed response.",
      "family": "testimonial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second still samples and unverified ASR; no playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 18,
          "role": "hook",
          "summary": "Speaker introduces enthusiasm and company context.",
          "visual": "Interview, street exterior, event signage and identification.",
          "evidence_time": 15.5,
          "mechanics": "Location samples widen the opening interview's setting.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.16
          }
        },
        {
          "start": 18,
          "end": 40,
          "role": "context",
          "summary": "Partnership endorsement meets an unpredictable future.",
          "visual": "Panel seating, interview closeup and attendee conversation.",
          "evidence_time": 30.5,
          "mechanics": "Public and informal settings surround direct testimony.",
          "asr_anchor": {
            "word_index": 94,
            "source_time": 19.0
          }
        },
        {
          "start": 40,
          "end": 65,
          "role": "qualification",
          "summary": "Speaker stresses large organizations' change burden.",
          "visual": "Panel audience, interview and mingling attendees.",
          "evidence_time": 55.5,
          "mechanics": "Repeated interview framing anchors the organizational caveat.",
          "asr_anchor": {
            "word_index": 294,
            "source_time": 47.1
          }
        },
        {
          "start": 65,
          "end": 80.974,
          "role": "close",
          "summary": "Speaker recommends reconciling desired and likely futures.",
          "visual": "Interview, conversation cutaway and closing logo.",
          "evidence_time": 75.5,
          "mechanics": "Direct testimony returns before the logo sample.",
          "asr_anchor": {
            "word_index": 470,
            "source_time": 68.0
          }
        }
      ],
      "editing_lesson": {
        "start": 0,
        "end": 36,
        "observation": "Street, panel and direct interview offer progressively narrower social context.",
        "inference": "Interpretation: the event situates the endorsement among practitioners.",
        "adaptation": "Film an original venue, discussion and participant reflection.",
        "caution": "Event attendance is not evidence of legal-product performance.",
        "evidence_times": [
          5.5,
          20.5,
          35.5
        ]
      },
      "codes": [
        {
          "code": "presenter-return",
          "beat_indices": [
            0,
            1,
            2,
            3
          ]
        }
      ],
      "uncertainty": "ASR inconsistently renders the company name; no legal workflow is demonstrated.",
      "media": {
        "sha256": "8ba304d0119950592b4e62ace5c1db6a27f8d25ee8a8f3be86105ea0c4962e5f",
        "duration": 80.974,
        "width": 854,
        "height": 480,
        "fps": "25/1",
        "retrieved_at": "2026-09-17T18:16:23.667043+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 280,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "8ba304d0119950592b4e62ace5c1db6a27f8d25ee8a8f3be86105ea0c4962e5f"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "jyuyY86GJnA",
      "rank": 57,
      "title": "Build Hour: Valuemaxxing with GPT-5.6",
      "source_url": "https://www.youtube.com/watch?v=jyuyY86GJnA",
      "publication_date": "2026-07-24",
      "publication_timestamp": "2026-07-24T12:47:16-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Speakers ask whether greater AI consumption produces useful work. Model comparisons, experimental results and a customer engineering account develop the question before audience discussion qualifies the trade-offs.",
      "family": "talk",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg",
          "sheet-02.jpg"
        ],
        "transcript_read": true,
        "basis": "Sixty-second samples, three enlarged sampled frames and unverified ASR; no playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 420,
          "role": "context",
          "summary": "Hosts contrast consumption metrics with useful outcomes.",
          "visual": "Two hosts, meme, agenda and four-part evaluation slide.",
          "evidence_time": 300.5,
          "mechanics": "Host-led layout becomes slide-dominant with a speaker inset.",
          "asr_anchor": {
            "word_index": 1064,
            "source_time": 170.1
          }
        },
        {
          "start": 420,
          "end": 915,
          "role": "proof",
          "summary": "Model comparisons lead to developer-use suggestions.",
          "visual": "Model cards, scatterplot, illustration grid and settings slide.",
          "evidence_time": 720.5,
          "mechanics": "One enlarged panda illustration supplements the output matrix.",
          "asr_anchor": {
            "word_index": 3890,
            "source_time": 580.004
          }
        },
        {
          "start": 915,
          "end": 1385,
          "role": "proof",
          "summary": "Speaker compares application-level efficiency techniques and reported results.",
          "visual": "Feature list followed by browser comparison tables.",
          "evidence_time": 1260.5,
          "mechanics": "Explanatory slides precede sampled experimental result panels.",
          "asr_anchor": {
            "word_index": 6088,
            "source_time": 930.08
          }
        },
        {
          "start": 1385,
          "end": 1740,
          "role": "context",
          "summary": "Customer guests introduce their marketing platform and engineering remit.",
          "visual": "Guest chapter card, portraits and website-workflow slides.",
          "evidence_time": 1620.5,
          "mechanics": "Yellow guest slides replace the earlier laboratory interface.",
          "asr_anchor": {
            "word_index": 9524,
            "source_time": 1470.368
          }
        },
        {
          "start": 1740,
          "end": 2385,
          "role": "proof",
          "summary": "Engineer explains caching, batching and smaller tool outputs.",
          "visual": "Colored context diagrams, code blocks and claimed savings.",
          "evidence_time": 1860.5,
          "mechanics": "Repeated column layouts compare retained and invalidated cache states.",
          "asr_anchor": {
            "word_index": 11060,
            "source_time": 1768.412
          }
        },
        {
          "start": 2385,
          "end": 3200,
          "role": "discussion",
          "summary": "Questions expose delegation, context and caching trade-offs.",
          "visual": "Audience-question cards beside hosts or remote guests.",
          "evidence_time": 2880.5,
          "mechanics": "Persistent question panels accompany changing speaker views.",
          "asr_anchor": {
            "word_index": 14632,
            "source_time": 2500.048
          }
        },
        {
          "start": 3200,
          "end": 3272.714,
          "role": "close",
          "summary": "Hosts share resources and announce another session.",
          "visual": "Two hosts beside a resource list.",
          "evidence_time": 3240.5,
          "mechanics": "Resource layout replaces the earlier question-card format.",
          "asr_anchor": {
            "word_index": 18772,
            "source_time": 3205.44
          }
        }
      ],
      "editing_lesson": {
        "start": 1740,
        "end": 2040,
        "observation": "Colored columns explain reuse, then invalidation; a later diagram adds code.",
        "inference": "Interpretation: benefit, failure and remedy make the optimization causal.",
        "adaptation": "Diagram an original workload, perturb one input, then demonstrate a repair.",
        "caution": "Reported savings concern the speakers' workloads, not universal performance.",
        "evidence_times": [
          1800.5,
          1860.5,
          1920.5
        ]
      },
      "codes": [
        {
          "code": "context-before-detail",
          "beat_indices": [
            0,
            1,
            2
          ]
        },
        {
          "code": "chapter-reset",
          "beat_indices": [
            1,
            3,
            5
          ]
        },
        {
          "code": "question-led",
          "beat_indices": [
            5
          ]
        }
      ],
      "uncertainty": "Minute gaps hide execution; ASR contains inconsistent names. The final 32 seconds lack a still sample.",
      "media": {
        "sha256": "a5b954dc4924f6b3d5b90e5791b5681d377f328ea07306d9f0365f1ec6938a18",
        "duration": 3272.714,
        "width": 854,
        "height": 480,
        "fps": "30/1",
        "retrieved_at": "2026-09-17T18:16:42.408866+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 9562,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "a5b954dc4924f6b3d5b90e5791b5681d377f328ea07306d9f0365f1ec6938a18"
      },
      "sampling_interval": 60,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "PB7DBemxl3E",
      "rank": 58,
      "title": "Building at the pace of AI innovation | Grant Lee, Gamma",
      "source_url": "https://www.youtube.com/watch?v=PB7DBemxl3E",
      "publication_date": "2026-07-24",
      "publication_timestamp": "2026-07-24T02:12:21-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "The speaker connects easier visual communication with rapidly changing AI capabilities. Partnership and internal-tool testimony develop into advice about adaptable teams, but sampled event footage does not establish the benefits.",
      "family": "testimonial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second still samples and unverified ASR; no playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 23,
          "role": "hook",
          "summary": "Speaker links company purpose to AI partnership.",
          "visual": "Gray-shirted interviewee and two-person event panel.",
          "evidence_time": 15.5,
          "mechanics": "Interview and panel framings establish two speaking contexts.",
          "asr_anchor": {
            "word_index": 0,
            "source_time": 0.04
          }
        },
        {
          "start": 23,
          "end": 57,
          "role": "context",
          "summary": "Speaker attributes faster development to tools and feedback.",
          "visual": "Event signage, panel closeup and interview gestures.",
          "evidence_time": 45.5,
          "mechanics": "Direct interview returns after event-context samples.",
          "asr_anchor": {
            "word_index": 182,
            "source_time": 25.12
          }
        },
        {
          "start": 57,
          "end": 76,
          "role": "discussion",
          "summary": "Speaker recommends collaborative, adaptable decision-making teams.",
          "visual": "Attendee conversations bracket an angled interview.",
          "evidence_time": 65.5,
          "mechanics": "Different conversational groupings surround the advice.",
          "asr_anchor": {
            "word_index": 434,
            "source_time": 60.24
          }
        },
        {
          "start": 76,
          "end": 90.614,
          "role": "close",
          "summary": "Speaker urges translating capabilities into customer benefits.",
          "visual": "Smiling attendee, interview and logo.",
          "evidence_time": 85.5,
          "mechanics": "Interview returns before the final identity card.",
          "asr_anchor": {
            "word_index": 552,
            "source_time": 78.14
          }
        }
      ],
      "editing_lesson": {
        "start": 35,
        "end": 76,
        "observation": "Interview gestures, a two-person conversation and an angled interview vary social scale.",
        "inference": "Interpretation: collaboration advice gains a visible human setting.",
        "adaptation": "Pair an original teamwork account with actual colleague interaction, then return to reflection.",
        "caution": "Networking footage cannot substantiate development-speed claims.",
        "evidence_times": [
          45.5,
          60.5,
          65.5
        ]
      },
      "codes": [
        {
          "code": "presenter-return",
          "beat_indices": [
            1,
            2,
            3
          ]
        }
      ],
      "uncertainty": "The sampled event scenes show neither internal coding nor slide generation.",
      "media": {
        "sha256": "25ef8c6da5477c2a36b9057ec27b41de158864f1204a6249bab520456f3bd33b",
        "duration": 90.614,
        "width": 854,
        "height": 480,
        "fps": "25/1",
        "retrieved_at": "2026-09-17T18:16:24.861011+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 311,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "25ef8c6da5477c2a36b9057ec27b41de158864f1204a6249bab520456f3bd33b"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "-_p2a6uWfjk",
      "rank": 59,
      "title": "What is AI Slop?",
      "source_url": "https://www.youtube.com/watch?v=-_p2a6uWfjk",
      "publication_date": "2026-07-23",
      "publication_timestamp": "2026-07-23T13:51:09-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "shorts",
        "uploads"
      ],
      "narrative": "A gallery visit asks who decides whether AI-associated art lacks value. Organizer and artist accounts meet tactile exhibits and visitor judgments, leaving the category contested rather than defined.",
      "family": "interview",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second portrait still samples and unverified ASR; no playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 22,
          "role": "hook",
          "summary": "Gallery introduction frames a question about artistic judgment.",
          "visual": "Question overlay, gallery signage and gathered visitors.",
          "evidence_time": 15.5,
          "mechanics": "Portrait text overlays locate the inquiry in an exhibition.",
          "asr_anchor": {
            "word_index": 22,
            "source_time": 7.2
          }
        },
        {
          "start": 22,
          "end": 42,
          "role": "context",
          "summary": "Participants describe organizational and creative assistance.",
          "visual": "Honeycomb lights, crowded room and display monitors.",
          "evidence_time": 35.5,
          "mechanics": "Different exhibit surfaces accompany the production account.",
          "asr_anchor": {
            "word_index": 122,
            "source_time": 24.04
          }
        },
        {
          "start": 42,
          "end": 63,
          "role": "discussion",
          "summary": "Artist explanation gives way to interactive judgments.",
          "visual": "Interviewee, keyboard hands, scratch-card hands and cartoon figure.",
          "evidence_time": 55.5,
          "mechanics": "Face-level testimony contrasts with close views of participation.",
          "asr_anchor": {
            "word_index": 252,
            "source_time": 43.06
          }
        },
        {
          "start": 63,
          "end": 81.634,
          "role": "close",
          "summary": "Participants leave quality judgments open to reinterpretation.",
          "visual": "Drawing-machine sign, game display, participant and closing card.",
          "evidence_time": 70.5,
          "mechanics": "Object and participant samples precede a minimal ending.",
          "asr_anchor": {
            "word_index": 406,
            "source_time": 65.04
          }
        }
      ],
      "editing_lesson": {
        "start": 42,
        "end": 63,
        "observation": "An interview face, keyboard use and scratch-card handling show different participation modes.",
        "inference": "Interpretation: judgment becomes a visible activity.",
        "adaptation": "Ask an original exhibition question, show visitors testing works, then collect reactions.",
        "caution": "Samples do not securely match every artist to an exhibit.",
        "evidence_times": [
          45.5,
          50.5,
          55.5
        ]
      },
      "codes": [
        {
          "code": "question-led",
          "beat_indices": [
            0,
            2,
            3
          ]
        }
      ],
      "uncertainty": "ASR names are unverified; sparse samples omit participant handoffs.",
      "media": {
        "sha256": "0c2b32631226fb73a570dc75fcf22d10046cc85354043a73c0a78c4210c49c47",
        "duration": 81.634,
        "width": 240,
        "height": 426,
        "fps": "30000/1001",
        "retrieved_at": "2026-09-17T18:16:25.089218+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 247,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "0c2b32631226fb73a570dc75fcf22d10046cc85354043a73c0a78c4210c49c47"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    },
    {
      "id": "t8Ej9aeW388",
      "rank": 60,
      "title": "How IT Admins can manage ChatGPT Work at scale | OpenAI",
      "source_url": "https://www.youtube.com/watch?v=t8Ej9aeW388",
      "publication_date": "2026-07-23",
      "publication_timestamp": "2026-07-23T13:50:48-07:00",
      "publication_date_source": "playerMicroformatRenderer.publishDate",
      "tabs": [
        "videos",
        "uploads"
      ],
      "narrative": "Two presenters frame workspace administration as understanding usage while retaining control. Scoped access leads to charts, a proposed spending cap and explicit confirmation, with final verification described more fully than sampled.",
      "family": "tutorial",
      "review": {
        "viewed_sheets": [
          "sheet-01.jpg"
        ],
        "transcript_read": true,
        "basis": "Five-second still samples and unverified ASR; no playback review."
      },
      "beats": [
        {
          "start": 0,
          "end": 32,
          "role": "context",
          "summary": "Presenters introduce administration and scoped access setup.",
          "visual": "Two presenters, closeup and key-creation modal.",
          "evidence_time": 25.5,
          "mechanics": "Screen insert interrupts repeated shared-desk framing.",
          "asr_anchor": {
            "word_index": 16,
            "source_time": 4.64
          }
        },
        {
          "start": 32,
          "end": 62,
          "role": "proof",
          "summary": "Usage questions produce activity and spending comparisons.",
          "visual": "Dark chat contains blue and orange bars.",
          "evidence_time": 55.5,
          "mechanics": "Activity charts precede differently ranked spending bars.",
          "asr_anchor": {
            "word_index": 186,
            "source_time": 33.26
          }
        },
        {
          "start": 62,
          "end": 76,
          "role": "qualification",
          "summary": "Proposed cap remains subject to review.",
          "visual": "Recommendation text and projected-impact slider.",
          "evidence_time": 75.5,
          "mechanics": "Text proposal gains comparative bars before acceptance.",
          "asr_anchor": {
            "word_index": 320,
            "source_time": 64.319
          }
        },
        {
          "start": 76,
          "end": 99.994,
          "role": "close",
          "summary": "Confirmation precedes the narrated verification and recap.",
          "visual": "Approval request in chat; both presenters return.",
          "evidence_time": 80.5,
          "mechanics": "Working-state interface gives way to shared-desk explanation.",
          "asr_anchor": {
            "word_index": 390,
            "source_time": 77.02
          }
        }
      ],
      "editing_lesson": {
        "start": 62,
        "end": 84,
        "observation": "Recommendation, impact preview and application request appear as separate states.",
        "inference": "Interpretation: preview makes human control legible.",
        "adaptation": "Record an original proposal, inspect its impact, approve and verify the changed setting.",
        "caution": "The sampled sequence does not show the final admin verification.",
        "evidence_times": [
          70.5,
          75.5,
          80.5
        ]
      },
      "codes": [
        {
          "code": "permission-step",
          "beat_indices": [
            2,
            3
          ]
        },
        {
          "code": "presenter-return",
          "beat_indices": [
            0,
            1,
            3
          ]
        }
      ],
      "uncertainty": "Security and cap-application claims are not validated by the sampled interface states.",
      "media": {
        "sha256": "b9f62ccd2405c12282b2486c610c9ba6522b350107552795a33fc0a50fdb419b",
        "duration": 99.994,
        "width": 854,
        "height": 480,
        "fps": "24000/1001",
        "retrieved_at": "2026-09-17T18:16:25.135322+00:00",
        "audio_present": true
      },
      "transcript": {
        "kind": "machine-asr",
        "provider": "ElevenLabs",
        "model": "scribe_v2",
        "status": "asr-unverified",
        "source_offset_seconds": 0.0,
        "speech_word_count": 260,
        "full_text_published": false,
        "quoted_words": 0,
        "timestamp_basis": "Returned ASR word starts plus measured stream offset; no human alignment certification.",
        "source_sha256": "b9f62ccd2405c12282b2486c610c9ba6522b350107552795a33fc0a50fdb419b"
      },
      "sampling_interval": 5,
      "independent_review": {
        "status": "pass",
        "scope": "Independent source-grounding against timed ASR and actual source images; not a continuous viewing/listening pass or validation of speaker claims.",
        "reviewed_on": "2026-09-17"
      }
    }
  ],
  "synthesis": {
    "presenter-return": {
      "label": "Return to the person",
      "definition": "An on-camera explanation returns after a screen or example. The return can close one task before the next begins.",
      "count": 22,
      "cases": [
        {
          "id": "B-jjnydci50",
          "beat_indices": [
            2,
            4
          ]
        },
        {
          "id": "bOC3DisEOfg",
          "beat_indices": [
            1,
            4
          ]
        },
        {
          "id": "NSqlZ1L81nQ",
          "beat_indices": [
            3
          ]
        },
        {
          "id": "-TTyyY3VWh8",
          "beat_indices": [
            1,
            3
          ]
        },
        {
          "id": "yfnh8SuMbNw",
          "beat_indices": [
            2,
            3
          ]
        },
        {
          "id": "syML5KT-HzI",
          "beat_indices": [
            4
          ]
        },
        {
          "id": "5QPBVrAAdBk",
          "beat_indices": [
            3
          ]
        },
        {
          "id": "0qK3KqvtRrg",
          "beat_indices": [
            2,
            3
          ]
        },
        {
          "id": "29SyCndnMZs",
          "beat_indices": [
            0,
            4
          ]
        },
        {
          "id": "Is2NHa7awWY",
          "beat_indices": [
            2
          ]
        },
        {
          "id": "jzmNh8lbSp8",
          "beat_indices": [
            3
          ]
        },
        {
          "id": "Cxp2aZQqkYM",
          "beat_indices": [
            0,
            1,
            2
          ]
        },
        {
          "id": "EDWbX_-RYws",
          "beat_indices": [
            1,
            2,
            4
          ]
        },
        {
          "id": "0XPk_MAwCW4",
          "beat_indices": [
            0,
            5
          ]
        },
        {
          "id": "W-HhMUe9hOg",
          "beat_indices": [
            1,
            2,
            4
          ]
        },
        {
          "id": "_HCks5jkPLw",
          "beat_indices": [
            0,
            2,
            3
          ]
        },
        {
          "id": "OEkxKdhtQng",
          "beat_indices": [
            3
          ]
        },
        {
          "id": "UaeWJK_vv-Y",
          "beat_indices": [
            2,
            3
          ]
        },
        {
          "id": "WeP9VUf1OoE",
          "beat_indices": [
            0,
            2,
            3,
            4
          ]
        },
        {
          "id": "oXu6JGxW_YM",
          "beat_indices": [
            0,
            1,
            2,
            3
          ]
        },
        {
          "id": "PB7DBemxl3E",
          "beat_indices": [
            1,
            2,
            3
          ]
        },
        {
          "id": "t8Ej9aeW388",
          "beat_indices": [
            0,
            1,
            3
          ]
        }
      ]
    },
    "input-to-artifact": {
      "label": "Keep the input and the deliverable",
      "definition": "A visible input is followed by an inspectable output. The edited sequence demonstrates a relationship, not real-time processing speed.",
      "count": 31,
      "cases": [
        {
          "id": "bOC3DisEOfg",
          "beat_indices": [
            1,
            3
          ]
        },
        {
          "id": "D-QteDHdHes",
          "beat_indices": [
            1,
            2,
            3,
            4
          ]
        },
        {
          "id": "Ez-anO32D_s",
          "beat_indices": [
            1,
            2,
            3
          ]
        },
        {
          "id": "N-MJ1W8Vj9E",
          "beat_indices": [
            1,
            3,
            4
          ]
        },
        {
          "id": "HHt8CPJRviA",
          "beat_indices": [
            1,
            2,
            3
          ]
        },
        {
          "id": "1QNsdr-Qx_I",
          "beat_indices": [
            0,
            1,
            4
          ]
        },
        {
          "id": "yfnh8SuMbNw",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "id": "syML5KT-HzI",
          "beat_indices": [
            1,
            2,
            3
          ]
        },
        {
          "id": "DLcQSp_CIIc",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "id": "urXc4xxixVU",
          "beat_indices": [
            1
          ]
        },
        {
          "id": "5QPBVrAAdBk",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "id": "0qK3KqvtRrg",
          "beat_indices": [
            1
          ]
        },
        {
          "id": "Rk4VyQxDq5s",
          "beat_indices": [
            2,
            3
          ]
        },
        {
          "id": "Lj03DNImB3s",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "id": "t64oZKCdG8Q",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "id": "Is2NHa7awWY",
          "beat_indices": [
            1
          ]
        },
        {
          "id": "jzmNh8lbSp8",
          "beat_indices": [
            1
          ]
        },
        {
          "id": "ceBruD6v5Bk",
          "beat_indices": [
            1,
            2,
            3
          ]
        },
        {
          "id": "Cxp2aZQqkYM",
          "beat_indices": [
            1
          ]
        },
        {
          "id": "bl2eGejkbaU",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "id": "94Td-y5z5rk",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "id": "mOAHWk24D28",
          "beat_indices": [
            2,
            3
          ]
        },
        {
          "id": "6VwIXjFj6YQ",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "id": "OEkxKdhtQng",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "id": "p_slDAvPjv0",
          "beat_indices": [
            2
          ]
        },
        {
          "id": "TryfaGZwvIE",
          "beat_indices": [
            1,
            2,
            3
          ]
        },
        {
          "id": "r64krUavXJU",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "id": "_Gd9yzAc-WI",
          "beat_indices": [
            0,
            4
          ]
        },
        {
          "id": "g-g6E-qtCjA",
          "beat_indices": [
            2,
            3
          ]
        },
        {
          "id": "5DbZAMGZ5oM",
          "beat_indices": [
            0,
            1,
            2
          ]
        },
        {
          "id": "hjXi7b01s2c",
          "beat_indices": [
            0,
            1,
            2
          ]
        }
      ]
    },
    "context-before-detail": {
      "label": "Locate the detail first",
      "definition": "An establishing view precedes a more specific state. Compare the sampled framing; do not assume a continuous zoom between them.",
      "count": 14,
      "cases": [
        {
          "id": "5iu0nY9lIy8",
          "beat_indices": [
            0,
            2
          ]
        },
        {
          "id": "HHt8CPJRviA",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "id": "NSqlZ1L81nQ",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "id": "urXc4xxixVU",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "id": "Rk4VyQxDq5s",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "id": "9MIkJHcMtVA",
          "beat_indices": [
            0
          ]
        },
        {
          "id": "0j9yDUDMrBs",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "id": "wIf9yMGalPU",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "id": "Ljm3nHQQiSk",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "id": "-vvGAKtV_Ek",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "id": "UaeWJK_vv-Y",
          "beat_indices": [
            0,
            1,
            2
          ]
        },
        {
          "id": "g-g6E-qtCjA",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "id": "q9TPE7pKkIo",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "id": "jyuyY86GJnA",
          "beat_indices": [
            0,
            1,
            2
          ]
        }
      ]
    },
    "output-montage": {
      "label": "Accumulate examples",
      "definition": "Multiple outputs carry the argument through variety. A montage is weaker evidence of the steps needed to make any one result.",
      "count": 7,
      "cases": [
        {
          "id": "bOC3DisEOfg",
          "beat_indices": [
            2
          ]
        },
        {
          "id": "D-QteDHdHes",
          "beat_indices": [
            2,
            4
          ]
        },
        {
          "id": "-TTyyY3VWh8",
          "beat_indices": [
            0,
            1,
            2
          ]
        },
        {
          "id": "1QNsdr-Qx_I",
          "beat_indices": [
            1,
            2,
            3
          ]
        },
        {
          "id": "2f9bBlGOuX4",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "id": "6VwIXjFj6YQ",
          "beat_indices": [
            3
          ]
        },
        {
          "id": "hjXi7b01s2c",
          "beat_indices": [
            2
          ]
        }
      ]
    },
    "human-stakes": {
      "label": "Give the task a human consequence",
      "definition": "A person, team or situation establishes why the work matters. The film’s success claims remain attributed claims.",
      "count": 5,
      "cases": [
        {
          "id": "Lj03DNImB3s",
          "beat_indices": [
            0,
            2,
            3
          ]
        },
        {
          "id": "t64oZKCdG8Q",
          "beat_indices": [
            0,
            3
          ]
        },
        {
          "id": "EDWbX_-RYws",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "id": "SJJBPwlPpww",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "id": "MLehRytu9Zo",
          "beat_indices": [
            0,
            3,
            4
          ]
        }
      ]
    },
    "question-led": {
      "label": "Let a question organize the explanation",
      "definition": "An explicit question or problem structures the following discussion or demonstration.",
      "count": 9,
      "cases": [
        {
          "id": "B-jjnydci50",
          "beat_indices": [
            0
          ]
        },
        {
          "id": "5iu0nY9lIy8",
          "beat_indices": [
            2,
            3
          ]
        },
        {
          "id": "-TTyyY3VWh8",
          "beat_indices": [
            0
          ]
        },
        {
          "id": "o38xYi2mtgc",
          "beat_indices": [
            1,
            4
          ]
        },
        {
          "id": "XjSJ6ybS9I8",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "id": "Ljm3nHQQiSk",
          "beat_indices": [
            0,
            1
          ]
        },
        {
          "id": "SJJBPwlPpww",
          "beat_indices": [
            1,
            3,
            6
          ]
        },
        {
          "id": "jyuyY86GJnA",
          "beat_indices": [
            5
          ]
        },
        {
          "id": "-_p2a6uWfjk",
          "beat_indices": [
            0,
            2,
            3
          ]
        }
      ]
    },
    "chapter-reset": {
      "label": "Separate changes of subject",
      "definition": "A title or explicit section reset distinguishes a new idea from the preceding example.",
      "count": 5,
      "cases": [
        {
          "id": "XjSJ6ybS9I8",
          "beat_indices": [
            1,
            2
          ]
        },
        {
          "id": "0j9yDUDMrBs",
          "beat_indices": [
            1
          ]
        },
        {
          "id": "yQZgOSHHxjk",
          "beat_indices": [
            1,
            3
          ]
        },
        {
          "id": "p_slDAvPjv0",
          "beat_indices": [
            1
          ]
        },
        {
          "id": "jyuyY86GJnA",
          "beat_indices": [
            1,
            3,
            5
          ]
        }
      ]
    },
    "permission-step": {
      "label": "Keep permission in the story",
      "definition": "A shown authorization or access decision remains part of the demonstrated causal chain.",
      "count": 8,
      "cases": [
        {
          "id": "981SivztzOc",
          "beat_indices": [
            2,
            3
          ]
        },
        {
          "id": "Rk4VyQxDq5s",
          "beat_indices": [
            1,
            3
          ]
        },
        {
          "id": "29SyCndnMZs",
          "beat_indices": [
            3
          ]
        },
        {
          "id": "W-HhMUe9hOg",
          "beat_indices": [
            0,
            3
          ]
        },
        {
          "id": "TryfaGZwvIE",
          "beat_indices": [
            3
          ]
        },
        {
          "id": "_Gd9yzAc-WI",
          "beat_indices": [
            3,
            4
          ]
        },
        {
          "id": "q9TPE7pKkIo",
          "beat_indices": [
            2
          ]
        },
        {
          "id": "t8Ej9aeW388",
          "beat_indices": [
            2,
            3
          ]
        }
      ]
    },
    "proof-before-close": {
      "label": "Finish with something inspectable",
      "definition": "An output, result or demonstrated state is shown before the closing passage. This does not independently verify a product claim.",
      "count": 4,
      "cases": [
        {
          "id": "DLcQSp_CIIc",
          "beat_indices": [
            2,
            3
          ]
        },
        {
          "id": "981SivztzOc",
          "beat_indices": [
            3,
            4
          ]
        },
        {
          "id": "bl2eGejkbaU",
          "beat_indices": [
            3
          ]
        },
        {
          "id": "WeP9VUf1OoE",
          "beat_indices": [
            3,
            4
          ]
        }
      ]
    }
  }
}