{ "schema_version": 1, "review_date": "2026-09-19", "route": "/videos/storytelling/script-writer/tonality/", "cohort_frozen_at": "2026-09-18T18:12:29.811865+00:00", "cohort_films": 100, "primary_texts_read": 25, "primary_machine_transcripts": 12, "additional_captioned_films": 13, "overlap_captioned_films": 8, "films_without_text_reading": 75, "listening_verified_films": 0, "new_source_media_downloads": 0, "classification": "Editorial groupings of an uneven text sample; films can mix registers.", "notes": "Original analysis. Fictional examples are not source quotations or verified product behavior. Passage times locate approximate source spans, not edit boundaries.", "modes": [ { "id": "guide", "label": "Patient teacher", "purpose": "Help someone take a small first step.", "description": "The beginner lessons reduce the cost of trying. They introduce one ordinary goal, allow an imperfect first attempt and give the learner a useful next action. Reassurance is attached to something the person can do.", "language": "Address the learner as you; use let's for a shared start. Prefer familiar verbs, concrete alternatives and permission to revise. Use a direct instruction when the action needs to be unambiguous.", "example": "Let's make a checklist for Friday's review. Start with the three things the team needs to check. You can change the order once you see the list.", "effect": "One small task makes the invitation manageable. The final sentence makes revision ordinary instead of treating the first attempt as a test.", "watch": "Avoid explaining every exception before the learner has a foothold. Give a relevant boundary beside the action it qualifies.", "references": [ "4UTn7oWvSJc", "P-8sCH6syag", "WY_PHjGilVs" ] }, { "id": "demonstrate", "label": "Working demonstrator", "purpose": "Let the audience follow a decision as you make it.", "description": "The demos often connect a first-person intention to an artifact the audience can inspect. Preferences, tentative ideas and unanswered questions can stay in the language. The point is to make the presenter's reasoning followable.", "language": "Move between I for the choice and we for the inspection. Name the object before referring to it as this. Keep maybe when you are exploring; distinguish a finding from a possible explanation.", "example": "I need a checklist for Friday's review. Here's the draft. The owner column is empty, so I'll add a person for each item before I share it.", "effect": "The visible gap supplies a reason for the next action. This is a planned demo line: only use it if the matching draft and edit are actually shown.", "watch": "A request is not a completed result. Avoid narrating a successful edit, share or scheduled update before the evidence supports it.", "references": [ "GzCsKuDgr6Q", "wK0zWeVyLM4", "0NUm552iB3U" ] }, { "id": "explain", "label": "Technical peer", "purpose": "Make a mechanism understandable without losing its conditions.", "description": "The technical films begin with recognizable work or a reasonable objection, then explain the context, tools and human responsibilities needed to address it. A single task can carry quite a lot of technical detail.", "language": "Use because, if and before to expose the causal chain. Name the dependency before the benefit. An everyday analogy can introduce the problem; return to the actual mechanism afterward.", "example": "A useful review checklist needs the current release scope and an owner for each item. We'll check those inputs before we use the list to decide whether Friday's release is ready.", "effect": "The prerequisites earn the confidence of the final clause. The mechanism is useful because it serves a decision the audience already understands.", "watch": "Do not turn an analogy into a claim of equivalent capability, or remove a dependency to make the sentence sound more confident.", "references": [ "bbNTY8O0GTY", "cXUmFec2PNY", "2YHa1vhnmK0" ] }, { "id": "human", "label": "Personal story", "purpose": "Show why the work matters to a particular person.", "description": "These stories allow memory, vulnerability, ambition and a more expressive ending. The tool enters an existing life: a small team, a persistent project, a demanding job or a scientific hope.", "language": "Use details that belong to this person. Name the practical cost before widening into a feeling or ambition. Keep reported experience, future possibility and achieved outcome distinct.", "example": "I used to spend Thursday evening asking who still needed to check the release. Now I have one list to work through with the team. I want Friday's meeting to be about the decisions we still need to make.", "effect": "A recurring evening task gives the aspiration a concrete cause. This is a fictional character's example, not customer testimony or a measured improvement.", "watch": "Do not write an intimate anecdote for a real person and present it as their experience. Ask for the detail that is actually theirs.", "references": [ "PSnJBfa183Y", "uA5PiAXKGis", "mPAmXQnhv2E" ] }, { "id": "business", "label": "Business case", "purpose": "Connect a capability or opinion to a consequential piece of work.", "description": "Launch walkthroughs and testimonials can be emphatic. Their more useful passages give the praise an object: a task, a constraint, a particular result or a remaining bottleneck. Attribution carries much of their credibility.", "language": "State the work and its consequence. Keep personal assessments in the speaker's voice. Preserve the scope, baseline and uncertainty of any reported outcome; finish with the decision or next responsibility.", "example": "For Friday's review, the checklist gives each open item an owner and a decision to make. We'll judge whether it helps by checking what the team resolves in the meeting.", "effect": "The example names an operational use and an evaluation to perform. It makes no invented claim that the checklist has already saved time or improved delivery.", "watch": "The corpus contains sweeping praise. Treat that as a feature of the testimonial, not a reliability guarantee you can transfer to another product.", "references": [ "VMgEQ7ym9WU", "Q_MEaNdMpZw", "A4BUbpKdenc" ] }, { "id": "compact", "label": "Compact exchange", "purpose": "Supply just enough spoken context for the scene to carry the rest.", "description": "The two short examples use work already in progress as a premise. A version, a problem or a deadline can make a brief request intelligible. Evidence is thinner here, especially where the recognition captures very few words.", "language": "Name the current object, the required change and the constraint. Omit exposition only when the picture or established context supplies it. Brevity should preserve the request's meaning.", "example": "The checklist is ready. Add the owners before Friday's review.", "effect": "The short exchange relies on an established checklist and review. Without that context it needs an introduction, regardless of how concise it sounds.", "watch": "Sparse recognized speech does not establish silence, a measured pause or the full soundtrack. Use the original film to assess performance and sound.", "references": [ "ceBruD6v5Bk", "PbfXbDbCthA" ] } ], "mode_counts": { "explain": 4, "demonstrate": 5, "business": 5, "guide": 5, "human": 4, "compact": 2 }, "films": [ { "id": "bbNTY8O0GTY", "mode": "explain", "title": "Answer the trust question before naming the mechanism", "reading": "The explanation anticipates a reasonable objection, inventories the missing context, then assigns review and maintenance to people. Conditions make its confidence more specific.", "passages": [ { "start": 2.8, "end": 33, "note": "A question about trusting the analysis leads to the practical problem of scattered definitions." }, { "start": 48.2, "end": 96, "note": "A knowledgeable reviewer and continuing organizational effort remain part of the account." } ], "limit": "Product names are garbled in places in the machine transcript; this reading does not depend on their exact spelling.", "film_title": "Use ChatGPT Work to ground analysis in your semantic layer", "rank": 1, "duration_seconds": 112, "source_url": "https://www.youtube.com/watch?v=bbNTY8O0GTY", "source_basis": "machine_transcript", "caption_overlap": true, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/bbNTY8O0GTY/", "close_reading": "/videos/storytelling/close-readings/films/bbNTY8O0GTY/" }, { "id": "GzCsKuDgr6Q", "mode": "demonstrate", "title": "Make the viewer's attention easy to direct", "reading": "The language moves from a concrete dashboard request to named features and locations. First-person plural descriptions let the audience inspect the result alongside the presenter.", "passages": [ { "start": 3, "end": 22.8, "note": "A task request gives way to a tour of the chart and tabs, with repeated spatial references." }, { "start": 22.8, "end": 36, "note": "Refresh controls are explained as choices the viewer can make." } ], "limit": "The caption wording locates features; it does not measure how long the task took to complete.", "film_title": "Use ChatGPT Work to build dashboards from a business question", "rank": 2, "duration_seconds": 42, "source_url": "https://www.youtube.com/watch?v=GzCsKuDgr6Q", "source_basis": "automatic_captions", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/GzCsKuDgr6Q/" }, { "id": "Wb8MhkpQpX8", "mode": "demonstrate", "title": "Treat preference as a useful design input", "reading": "Hypothetical preferences make customization feel conversational. The speaker moves from labels and chart types to a reason for changing the view: different people need different readings of the same information.", "passages": [ { "start": 3.3, "end": 24, "note": "Possible terminology and chart preferences frame the edits as ordinary choices." }, { "start": 37.9, "end": 65, "note": "A more detailed exploration ends by returning to the stakeholder's way of understanding the business." } ], "limit": "Casual praise is part of this demonstration's language; restrained vocabulary is not a universal rule across the films.", "film_title": "Use ChatGPT Work to tailor dashboards for every stakeholder", "rank": 3, "duration_seconds": 70, "source_url": "https://www.youtube.com/watch?v=Wb8MhkpQpX8", "source_basis": "automatic_captions", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/Wb8MhkpQpX8/" }, { "id": "euED8JRunBA", "mode": "explain", "title": "Start from a problem the audience recognizes", "reading": "Addressing data-team leaders through their workload creates peer recognition. The account then places a trustworthy foundation before the wider promise of self-service.", "passages": [ { "start": 3.1, "end": 13, "note": "The opening identifies a familiar imbalance between available analysts and incoming questions." }, { "start": 25.1, "end": 38, "note": "The prerequisite comes before the description of broader participation." } ], "limit": "Reported internal adoption is the speaker's account, not an independent usage audit.", "film_title": "Use ChatGPT Work to help teams answer their own data questions", "rank": 4, "duration_seconds": 43, "source_url": "https://www.youtube.com/watch?v=euED8JRunBA", "source_basis": "automatic_captions", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/euED8JRunBA/" }, { "id": "wK0zWeVyLM4", "mode": "demonstrate", "title": "Make uncertainty part of competent reasoning", "reading": "An imagined leadership decision supplies a reason for the demo. The presenter distinguishes an observed drop-off from competing explanations, seeks more context, then owns the recommendation.", "passages": [ { "start": 3.5, "end": 16.9, "note": "A meeting and an investment decision give the analysis an immediate purpose." }, { "start": 65.4, "end": 110.2, "note": "Two possible explanations lead to a further inquiry before a recommendation is made." } ], "limit": "The film presents a demonstration scenario. Its example data and diagnosis should not be rewritten as verified results for another product.", "film_title": "Use ChatGPT Work to turn product insights into a team action plan", "rank": 5, "duration_seconds": 139, "source_url": "https://www.youtube.com/watch?v=wK0zWeVyLM4", "source_basis": "automatic_captions", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/wK0zWeVyLM4/" }, { "id": "cXUmFec2PNY", "mode": "explain", "title": "Use an analogy to explain what the system needs", "reading": "The new-analyst analogy translates context setup into a familiar human task. First-person actions then become shared instructions, with access conditions and human review still visible in the language.", "passages": [ { "start": 2.9, "end": 18.3, "note": "The speaker asks what a new analyst would need before describing tool access." }, { "start": 41.9, "end": 85.8, "note": "Conditional access requirements precede the explanation of assembling and reviewing shared context." } ], "limit": "The analogy explains the task; it does not establish equivalence between an agent and a human analyst.", "film_title": "Use ChatGPT Work to build your data agent", "rank": 6, "duration_seconds": 92, "source_url": "https://www.youtube.com/watch?v=cXUmFec2PNY", "source_basis": "machine_transcript", "caption_overlap": true, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/cXUmFec2PNY/", "close_reading": "/videos/storytelling/close-readings/films/cXUmFec2PNY/" }, { "id": "VMgEQ7ym9WU", "mode": "business", "title": "Carry a product promise through a business decision", "reading": "The polished walkthrough moves from a management question to evidence, a recommendation, owners and a shared view. Repeated returns to the team's next decision keep the feature sequence purposeful.", "passages": [ { "start": 12.4, "end": 16.1, "note": "The opening frames the promised journey as a question becoming an actionable plan." }, { "start": 46.7, "end": 85.4, "note": "Inspection, contextual inquiry and a reviewed handoff connect the analysis to responsibility." } ], "limit": "This is a scripted product case. Assertions of trust and consistency remain source claims, not independent validation.", "film_title": "Meet the Data Agent in ChatGPT Work", "rank": 7, "duration_seconds": 215, "source_url": "https://www.youtube.com/watch?v=VMgEQ7ym9WU", "source_basis": "automatic_captions", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/VMgEQ7ym9WU/" }, { "id": "s495saGYWFg", "mode": "business", "title": "Keep sweeping praise attached to its speaker", "reading": "The opening makes a broad capability claim; the account later adds carefully scoped capabilities and a description of demanding work. Its energy comes partly from strong testimonial language.", "passages": [ { "start": 0, "end": 6.7, "note": "A confident opening compresses a large claim about ambiguous tasks." }, { "start": 23.8, "end": 46.5, "note": "The explanation adds scope, information sources and the judgments the work requires." } ], "limit": "Personal enthusiasm is evidence of the testimonial's register, not a general reliability guarantee.", "film_title": "Jump Trading points GPT-6 Astra to its most ambiguous, difficult tasks", "rank": 10, "duration_seconds": 59, "source_url": "https://www.youtube.com/watch?v=s495saGYWFg", "source_basis": "automatic_captions", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/s495saGYWFg/" }, { "id": "Q_MEaNdMpZw", "mode": "business", "title": "Move between a large thesis and operational detail", "reading": "Role identification and a three-level framing establish a leadership perspective. Later alternatives, human escalation and qualified estimates give the broad thesis a more concrete operating context.", "passages": [ { "start": 7, "end": 39.4, "note": "The speaker identifies his role, frames change at several organizational levels and reports outcomes." }, { "start": 61.6, "end": 84.6, "note": "The help-desk account branches between self-service, human escalation and resolution." } ], "limit": "Reported outcomes and estimates belong to this account. Preserve attribution and qualifiers when discussing them.", "film_title": "Reimagining IT with ChatGPT", "rank": 11, "duration_seconds": 209, "source_url": "https://www.youtube.com/watch?v=Q_MEaNdMpZw", "source_basis": "machine_transcript", "caption_overlap": true, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/Q_MEaNdMpZw/", "close_reading": "/videos/storytelling/close-readings/films/Q_MEaNdMpZw/" }, { "id": "4UTn7oWvSJc", "mode": "guide", "title": "Make a small first attempt feel sufficient", "reading": "The instructor narrows the goal, normalizes an imperfect first answer and gives the learner choices. The reassurance becomes useful because it is attached to a repeatable action.", "passages": [ { "start": 5.6, "end": 22.4, "note": "A small shared goal is followed by permission to ask again." }, { "start": 47.9, "end": 85.1, "note": "Selection and revision remain the learner's choices; important answers require checking." } ], "limit": "The request for a simpler answer appears in the studied sequence, but a revised answer is not shown there. Avoid narrating an unseen result.", "film_title": "Meet ChatGPT: Ask Your First Question | OpenAI Academy", "rank": 12, "duration_seconds": 103, "source_url": "https://www.youtube.com/watch?v=4UTn7oWvSJc", "source_basis": "machine_transcript", "caption_overlap": true, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/4UTn7oWvSJc/", "close_reading": "/videos/storytelling/close-readings/films/4UTn7oWvSJc/" }, { "id": "P-8sCH6syag", "mode": "guide", "title": "Teach specificity without demanding polished prose", "reading": "The directions-to-a-friend analogy leads to practical details about audience, constraints and output. Revision is described as continuing a conversation, reducing the sense of a one-shot test.", "passages": [ { "start": 5.2, "end": 23.4, "note": "An imperfect answer is treated as a reason to clarify rather than stop." }, { "start": 64.1, "end": 85.9, "note": "The learner can revise part of the request; complete or elaborate sentences are not the central requirement." } ], "limit": "Caption line breaks do not establish spoken pauses or the instructor's exact cadence.", "film_title": "How to Ask ChatGPT Better Questions | OpenAI Academy", "rank": 13, "duration_seconds": 108, "source_url": "https://www.youtube.com/watch?v=P-8sCH6syag", "source_basis": "automatic_captions", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/P-8sCH6syag/" }, { "id": "gKyEzP_jhZc", "mode": "guide", "title": "End the help with a human decision", "reading": "A domestic analogy makes organization tangible. The script permits a shorter, more useful format and ends by returning judgment about the plan to the person using it.", "passages": [ { "start": 5.2, "end": 19.7, "note": "A messy pile becoming a short list gives the abstract benefit a familiar shape." }, { "start": 40.5, "end": 51.7, "note": "The speaker models asking for a more manageable format." }, { "start": 79.9, "end": 92, "note": "Checking details and deciding whether the plan fits remain human responsibilities." } ], "limit": "This is a text reading of agency and qualification, not a listening assessment of reassurance.", "film_title": "Use ChatGPT to Organize Everyday Tasks | OpenAI Academy", "rank": 14, "duration_seconds": 98, "source_url": "https://www.youtube.com/watch?v=gKyEzP_jhZc", "source_basis": "machine_transcript", "caption_overlap": true, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/gKyEzP_jhZc/", "close_reading": "/videos/storytelling/close-readings/films/gKyEzP_jhZc/" }, { "id": "WY_PHjGilVs", "mode": "guide", "title": "Use direct instructions when the action matters", "reading": "The script identifies pressure tactics, then replaces their urgency with specific steps. Firm imperative language serves the learner's decision; warmth does not require hedging every instruction.", "passages": [ { "start": 18.7, "end": 30.4, "note": "The explanation turns pressure to act quickly into a reason to pause." }, { "start": 63.5, "end": 81.7, "note": "Verification is described through concrete actions and an independent destination." } ], "limit": "This guide studies how the instruction is written; it is not a new review of scam-detection capability.", "film_title": "How to Spot Scam Messages with ChatGPT | OpenAI Academy", "rank": 15, "duration_seconds": 109, "source_url": "https://www.youtube.com/watch?v=WY_PHjGilVs", "source_basis": "automatic_captions", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/WY_PHjGilVs/" }, { "id": "dfAZbuAQxi8", "mode": "guide", "title": "Make a boundary usable with a concrete example", "reading": "A packing analogy and an ordinary bill make a general privacy principle easier to apply. Later clauses separate preparation and understanding from professional or official judgment.", "passages": [ { "start": 12, "end": 38.4, "note": "The analogy leads to deciding which details matter to the question." }, { "start": 66.1, "end": 95.4, "note": "The script names useful supporting tasks, then places the important check elsewhere." } ], "limit": "The analysis concerns rhetorical structure, not the current privacy settings or legal terms of a product.", "film_title": "ChatGPT Privacy: What Not to Share and When to Double-Check | OpenAI Academy", "rank": 16, "duration_seconds": 101, "source_url": "https://www.youtube.com/watch?v=dfAZbuAQxi8", "source_basis": "automatic_captions", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/dfAZbuAQxi8/" }, { "id": "PSnJBfa183Y", "mode": "human", "title": "Let the person's persistence carry the emotion", "reading": "Large stakes give way to a young inventor's history, field activity and personal motivation. Specific memories and a sense of belonging provide warmth without requiring every sentence to advertise the tool.", "passages": [ { "start": 28, "end": 60.4, "note": "A school project and years of improvement establish the person's history." }, { "start": 104.2, "end": 118.9, "note": "Urgency becomes a personal account of purpose and belonging." } ], "limit": "Potential benefit and testimony do not establish detection performance or lives saved.", "film_title": "Detecting wildfires with ChatGPT", "rank": 20, "duration_seconds": 122, "source_url": "https://www.youtube.com/watch?v=PSnJBfa183Y", "source_basis": "subtitles", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/PSnJBfa183Y/" }, { "id": "uA5PiAXKGis", "mode": "human", "title": "Connect a product benefit to a recognizable cost in someone's life", "reading": "The film opens with a striking customer anecdote, then explains the evening work behind it. Team details and accounts of improving drafts give the personal promise a working context.", "passages": [ { "start": 0, "end": 35.3, "note": "An attributed customer story becomes a concrete conflict between administrative work and family time." }, { "start": 89.6, "end": 124.1, "note": "The account shifts to an engineer's first assignment and the work involved in improving drafts." } ], "limit": "The customer anecdote and business claims are attributed testimony. Some model and company names are imperfectly captured in the automatic captions.", "film_title": "How Fyxer built an AI executive assistant people trust", "rank": 21, "duration_seconds": 196, "source_url": "https://www.youtube.com/watch?v=uA5PiAXKGis", "source_basis": "automatic_captions", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/uA5PiAXKGis/" }, { "id": "0NUm552iB3U", "mode": "demonstrate", "title": "Keep exploratory language when the task is exploration", "reading": "Approximate timing, first-person intentions and conditional visual ideas make this a designer's account of trying things. Naming adjustable properties gives the informal commentary something concrete to point to.", "passages": [ { "start": 0, "end": 8.4, "note": "The speaker describes an approximate personal building experience." }, { "start": 21.5, "end": 42.2, "note": "A possible letterform develops into named visual controls." } ], "limit": "The reported build time is approximate testimony. The transcript does not verify it, or prove which control caused every visual change.", "film_title": "GPT-6 Astra built a font playground", "rank": 26, "duration_seconds": 53, "source_url": "https://www.youtube.com/watch?v=0NUm552iB3U", "source_basis": "machine_transcript", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/0NUm552iB3U/", "close_reading": "/videos/storytelling/close-readings/films/0NUm552iB3U/" }, { "id": "2YHa1vhnmK0", "mode": "explain", "title": "Earn technical density with one coherent task", "reading": "An inventory of infrastructure work precedes the launch claim. A single incident-investigation example then carries the explanation through tools, context, execution and a report someone can review.", "passages": [ { "start": 0, "end": 23.8, "note": "The problem is broken into named responsibilities before the product is introduced." }, { "start": 28.3, "end": 115.5, "note": "Technical terms remain connected to the requirements and deliverable of one investigation." } ], "limit": "The workflow illustrates the explanation. Its token comparison and outcome are not independently reproduced benchmarks.", "film_title": "Introducing the Agents API", "rank": 27, "duration_seconds": 138, "source_url": "https://www.youtube.com/watch?v=2YHa1vhnmK0", "source_basis": "machine_transcript", "caption_overlap": true, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/2YHa1vhnmK0/", "close_reading": "/videos/storytelling/close-readings/films/2YHa1vhnmK0/" }, { "id": "PbfXbDbCthA", "mode": "compact", "title": "Leave room for the visual idea", "reading": "The retained recognition contains only a brief workplace request involving a project update, a launch deck and meeting notes. It offers a compact task premise for the separately studied visual sequence.", "passages": [ { "start": 0, "end": 6, "note": "The available spoken text supplies a work request instead of a long explanation." } ], "limit": "This is a very small, unverified ASR sample. Its exact opening syntax is uncertain. Lack of later recognized words does not establish silence, music or complete speech coverage.", "film_title": "In Harmony", "rank": 28, "duration_seconds": 30, "source_url": "https://www.youtube.com/watch?v=PbfXbDbCthA", "source_basis": "machine_transcript", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/PbfXbDbCthA/", "close_reading": "/videos/storytelling/close-readings/films/PbfXbDbCthA/" }, { "id": "mPAmXQnhv2E", "mode": "human", "title": "Make the capability gap personal and specific", "reading": "The account names the practical limits of a small family team and the roles they still have to fill. The ending widens into pride, freedom and ambition, allowing more expressive language than the instructional films.", "passages": [ { "start": 17.2, "end": 54.9, "note": "Tour history, team size and unfamiliar roles establish the need for help." }, { "start": 87.7, "end": 109.2, "note": "The payoff is expressed through family participation and entrepreneurial possibility." } ], "limit": "Some technical nouns are mistranscribed. The family account does not independently establish financial results or the correctness of advice shown in the film.", "film_title": "Helping a small business go big with ChatGPT", "rank": 30, "duration_seconds": 113, "source_url": "https://www.youtube.com/watch?v=mPAmXQnhv2E", "source_basis": "machine_transcript", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/mPAmXQnhv2E/", "close_reading": "/videos/storytelling/close-readings/films/mPAmXQnhv2E/" }, { "id": "0XLGisanEAs", "mode": "human", "title": "Preserve the difference between a possibility and an achieved benefit", "reading": "An imagined human benefit opens the story. The language moves through scientific work and vivid biological examples, then returns to a hoped-for outcome and personal motivation.", "passages": [ { "start": 0.1, "end": 19.4, "note": "The opening asks the viewer to imagine a consequence of the research." }, { "start": 110.1, "end": 120, "note": "The ending describes a desired benefit as a dream that motivates the work." } ], "limit": "Research claims and hopes in the film are not independently verified clinical outcomes. Keep that distinction when adapting the structure.", "film_title": "Discovering new antibiotics with ChatGPT", "rank": 31, "duration_seconds": 129, "source_url": "https://www.youtube.com/watch?v=0XLGisanEAs", "source_basis": "machine_transcript", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/0XLGisanEAs/", "close_reading": "/videos/storytelling/close-readings/films/0XLGisanEAs/" }, { "id": "n-dqdVMwoas", "mode": "business", "title": "Follow broad enthusiasm with a particular test", "reading": "The montage allows praise, surprise and personal ambition. Its more assessable passages name a task, a possible mistake or a bounded reported improvement, giving the enthusiasm an object.", "passages": [ { "start": 13.2, "end": 75, "note": "Different speakers describe their impressions and the work that interests them." }, { "start": 79.9, "end": 106.9, "note": "An example explains the consequence of a specific calculation error." }, { "start": 198.3, "end": 214.1, "note": "A reported optimization is tied to a particular workload." } ], "limit": "A stitched set of first impressions is not a controlled comparison or a universal performance guarantee.", "film_title": "GPT-6 Astra First Impressions From Businesses", "rank": 35, "duration_seconds": 228, "source_url": "https://www.youtube.com/watch?v=n-dqdVMwoas", "source_basis": "subtitles", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/n-dqdVMwoas/" }, { "id": "A4BUbpKdenc", "mode": "business", "title": "Let candid constraints qualify a strong opinion", "reading": "The speaker describes demanding tests, admits problems in his own code and reports a personal reliability improvement. The conversation ends with a remaining compute constraint, keeping the account situated in actual work.", "passages": [ { "start": 12.8, "end": 25.9, "note": "The speaker explains why he chooses difficult evaluation tasks." }, { "start": 99.7, "end": 152.7, "note": "A candid account of the existing code precedes praise, then a practical bottleneck." } ], "limit": "Descriptions of the model's communication style are the speaker's judgments about the model, not a listening analysis of the film's vocal delivery.", "film_title": "GPT-6 Astra with Peter Gostev", "rank": 42, "duration_seconds": 159, "source_url": "https://www.youtube.com/watch?v=A4BUbpKdenc", "source_basis": "subtitles", "caption_overlap": false, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/A4BUbpKdenc/" }, { "id": "0qK3KqvtRrg", "mode": "demonstrate", "title": "Use a continuing personal task to connect instructions", "reading": "The speaker connects an existing trip plan to a new artifact, then gives requirements, asks for review and requests changes. A later work example broadens the use without abandoning task-oriented language.", "passages": [ { "start": 11.8, "end": 46.3, "note": "The personal context leads into a concrete request and a review condition." }, { "start": 61.7, "end": 95.6, "note": "Revision and sharing are expressed as choices and requests." } ], "limit": "Requested sharing and future updates are not proof of recipient access or a later scheduled execution. Keep request, preview and confirmed result separate.", "film_title": "Build a Shareable Site", "rank": 59, "duration_seconds": 141, "source_url": "https://www.youtube.com/watch?v=0qK3KqvtRrg", "source_basis": "machine_transcript", "caption_overlap": true, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/0qK3KqvtRrg/", "close_reading": "/videos/storytelling/close-readings/films/0qK3KqvtRrg/" }, { "id": "ceBruD6v5Bk", "mode": "compact", "title": "Make a short exchange sound like work already in progress", "reading": "Brief clauses identify a version, a specific problem and a deadline. The request carries enough context for a response while leaving the visual sequence to explain the wider setting.", "passages": [ { "start": 5.9, "end": 18.4, "note": "Assessment, comparison and a deadline fit into a compact workplace exchange." }, { "start": 20, "end": 31, "note": "The available text moves from a comparison response into a concrete ingredient explanation." } ], "limit": "Ingredient names differ between machine sources. Speaker identity, intonation and conversational timing have not been checked by listening.", "film_title": "You can just use your voice | ChatGPT Work", "rank": 69, "duration_seconds": 40, "source_url": "https://www.youtube.com/watch?v=ceBruD6v5Bk", "source_basis": "machine_transcript", "caption_overlap": true, "listening_verified": false, "story_study": "/videos/storytelling/latest-100/films/ceBruD6v5Bk/", "close_reading": "/videos/storytelling/close-readings/films/ceBruD6v5Bk/" } ] }