{
  "docId": "019dd923-5e88-73ef-bd5d-0856e1444fb9",
  "docSlug": "8757f1b44ef7f176",
  "documentTitle": "2024 Air Street Capital The State of AI Report 2024",
  "authorId": "AirStreetCapital",
  "authorName": "Air Street Capital",
  "documentKindSlug": "consulting-deck",
  "documentKindLabel": "Consulting deck",
  "sourceTypeSlug": "vc_research",
  "sourceTypeLabel": "VC research",
  "presentationDate": null,
  "orientation": "landscape",
  "aspectRatio": 1.777,
  "pageNumber": 122,
  "pageCount": 213,
  "prevPage": 121,
  "nextPage": 123,
  "slideType": "industry_trends",
  "function": "summarize",
  "density": "overcrowded",
  "nDataPoints": 1,
  "notes": "Includes a screenshot of a tool used to check if YouTube videos were used for AI training.",
  "elementsJson": [
    "headline_text",
    "bullet_list",
    "screenshot"
  ],
  "metadataConfidence": 1,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5d-0856e1444fb9/122",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5d-0856e1444fb9",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5d-0856e1444fb9.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5d-0856e1444fb9#slide-122",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": "While copyright concerns are nothing new in generative AI, 2024 saw model builders come under greater scrutiny from media organizations, record labels, and content creators.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-47c3-7407-8d1b-a181b6c32cb8",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.45,
        "w": 0.27,
        "x": 0.71,
        "y": 0.435
      },
      "kind": "image",
      "text": "Screenshot of a tool checking YouTube video usage for AI training.",
      "attrs": null,
      "subkind": "screenshot",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "4e3ee0a2-f4db-41e8-8239-afcab8b28f13",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.08,
        "w": 0.68,
        "x": 0.02,
        "y": 0.44
      },
      "kind": "list",
      "text": "Some start-ups are swerving this altogether and are embracing ethical certification schemes. The best-known is Fairly Trained, started by ex-Stability AI executive Ed Newton-Rex.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "2f2c2b52-9b01-46ba-b911-c7e55b7c2ab2",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.1,
        "w": 0.68,
        "x": 0.02,
        "y": 0.33
      },
      "kind": "list",
      "text": "OpenAI and Google are negotiating with major media organizations, hoping that licensing arrangements will take the sting out of criticism. In a similar vein, Eleven Labs has started a voice actor program.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "41dfbb90-6ef0-4e4c-bd94-cef90b4279ef",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.18,
        "w": 0.68,
        "x": 0.02,
        "y": 0.62
      },
      "kind": "list",
      "text": "As labs approach the data ceiling, YouTube scraping is in the spotlight. OpenAI reportedly transcribed millions of hours of YouTube videos to power its audio transcription model. Meanwhile, Eleuther AI’s widely-used Pile dataset contains subtitles from 173,536 YouTube videos. Internal documents from both RunwayML and NVIDIA showed they mass scraped YouTube.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "7bf236f2-7626-484b-960f-712e2dc2cc9f",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.08,
        "w": 0.68,
        "x": 0.02,
        "y": 0.53
      },
      "kind": "list",
      "text": "At the other end of the spectrum, Meta and Perplexity have doubled down on 'fair use' arguments and have demonstrated little appetite for compromise with their critics.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "f7bc3a4d-8038-4d9e-84e4-359cb70af175",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": null,
      "kind": "quote",
      "text": "My videos in an AI training set called \"the pile\". I didn't authorize this and it goes against YouTube's ToS. Are all of the companies using this going to compensate creators of this source data? lol. - Scott J. Waldron (@PhotoBanzai)",
      "attrs": null,
      "subkind": null,
      "toolName": "Authority citation",
      "toolSlug": "authority-citation",
      "confidence": null,
      "componentId": "019dd952-47c3-7407-8d1b-a7d80132509b",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.05,
        "w": 0.65,
        "x": 0.02,
        "y": 0.14
      },
      "kind": "title",
      "text": "Industry attitudes to copyright diverge as anger from content creators rises...",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "7fdd86fb-97fd-405c-9208-8e1f52d054e9",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [
    {
      "name": "Action Titles",
      "slug": "action-titles",
      "agent": "Architect",
      "layer": "slide",
      "matchId": "019dd95a-1055-74e0-a959-e902b33668fc",
      "evidence": "Title 'Industry attitudes to copyright diverge as anger from content creators rises...'",
      "confidence": 82
    },
    {
      "name": "List presentation",
      "slug": "list-presentation",
      "agent": null,
      "layer": "slide",
      "matchId": "bb039113-4269-4e35-8321-9a546238de4b",
      "evidence": "The slide contains a list/bullet points with information about industry attitudes to copyright.",
      "confidence": 0.8
    }
  ],
  "frameworks": [],
  "arcBeats": [
    {
      "to": 153,
      "from": 9,
      "beatId": "019dd95a-0682-776c-8e35-5f2398b8d1d0",
      "arcName": "The Triple Take",
      "arcSlug": "triple-take",
      "beatName": "The Facts (What)",
      "beatSlug": "triple-take-the-facts-what",
      "evidence": "Exec summary + Research + Industry sections inventory the year",
      "position": 1,
      "confidence": 78,
      "parentBeatName": "Setup",
      "parentBeatSlug": "setup"
    },
    {
      "to": 174,
      "from": 86,
      "beatId": "019dd95a-0682-776c-8e35-73133c63a096",
      "arcName": "Voyage and Return",
      "arcSlug": "voyage-return",
      "beatName": "Discoveries",
      "beatSlug": "voyage-return-discoveries",
      "evidence": "Industry + Politics findings on NVIDIA, regulation, geopolitics",
      "position": 3,
      "confidence": 55,
      "parentBeatName": "Development",
      "parentBeatSlug": "development"
    }
  ],
  "loops": [
    {
      "to": 130,
      "from": 122,
      "name": "Pattern Hunter",
      "slug": "02-pattern-hunter",
      "bestFor": "Time-pressed audiences, building consensus, when data is strong",
      "matchId": "019dd95a-07fe-70ce-8d3e-555aebcfb636",
      "evidence": "Copyright suits, Wayve/Waymo, humanoids, Hollywood/Lionsgate, lab fragmentation",
      "position": 14,
      "objective": "Survey industry maneuvering: copyright, AVs, humanoids, talent",
      "structure": "Evidence A -> Evidence B -> Evidence C -> Pattern/Conclusion",
      "confidence": 65,
      "description": "Group multiple pieces of evidence that together point to a pattern or conclusion"
    }
  ],
  "imagePathAlt": null,
  "thumbSrc": null,
  "thumbSrcAlt": null,
  "locked": true
}