{
  "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": 40,
  "pageCount": 213,
  "prevPage": 39,
  "nextPage": 41,
  "slideType": "industry_trends",
  "function": "summarize",
  "density": "balanced",
  "nDataPoints": 0,
  "notes": "The slide uses a 'before-after' framing to highlight the rapid advancement in AI vision capabilities.",
  "elementsJson": [
    "headline_text",
    "paragraph",
    "screenshot",
    "other"
  ],
  "metadataConfidence": 0.95,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5d-0856e1444fb9/40",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5d-0856e1444fb9",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5d-0856e1444fb9.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5d-0856e1444fb9#slide-40",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": "Every major frontier model builder now offers vision capabilities out of the box.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-47c3-7407-8d11-b92a4ac35587",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.3,
        "w": 0.34,
        "x": 0.04,
        "y": 0.58
      },
      "kind": "image",
      "text": "2018 examples of AI vision errors",
      "attrs": null,
      "subkind": "photo",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "1d21d1f6-0989-469a-b07c-71cff75ab759",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.3,
        "w": 0.58,
        "x": 0.39,
        "y": 0.58
      },
      "kind": "image",
      "text": "2024 VLM performance demo",
      "attrs": null,
      "subkind": "screenshot",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "7c130a77-ce96-411a-bb91-f9b5473c2141",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.25,
        "w": 0.96,
        "x": 0.02,
        "y": 0.22
      },
      "kind": "paragraph",
      "text": "The first State of AI Report in 2018 detailed the painstaking efforts of researchers who tried to teach models common sense scene understanding by creating datasets of millions of labelled videos. Every major frontier model builder now offers vision capabilities out of the box. Even smaller models, in the low hundreds M to single digit B parameter size like Florence-2 from Microsoft or LongVILA from NVIDIA, can achieve remarkable results. Allen Institute for AI's open source Molmo can hold its own against the larger, proprietary GPT-4o.",
      "attrs": null,
      "subkind": "paragraph",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "feaa7775-00b2-48ed-a6eb-306b94d2b86a",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.05,
        "w": 0.5,
        "x": 0.02,
        "y": 0.14
      },
      "kind": "title",
      "text": "VLMs achieve SOTA performance out-of-the-box",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "39704db7-9ca2-4646-9fa6-809f71f3ae77",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [
    {
      "name": "Action Titles",
      "slug": "action-titles",
      "agent": "Architect",
      "layer": "slide",
      "matchId": "019dd95a-1055-74e0-a958-d3ca5901fe42",
      "evidence": "Title 'VLMs achieve SOTA performance out-of-the-box'",
      "confidence": 85
    }
  ],
  "frameworks": [
    {
      "name": "before-after-framing",
      "slug": null,
      "matchId": "0a19994d-1fdd-456e-b20f-ec04becc7eb6",
      "evidence": "The slide explicitly contrasts 2018 efforts with 2024 capabilities.",
      "confidence": 0.9
    }
  ],
  "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": 85,
      "from": 12,
      "beatId": "019dd95a-0682-776c-8e35-6d9d8ee06d0a",
      "arcName": "Voyage and Return",
      "arcSlug": "voyage-return",
      "beatName": "The Unknown",
      "beatSlug": "voyage-return-the-unknown",
      "evidence": "Research section explores frontier model uncharted territory",
      "position": 2,
      "confidence": 55,
      "parentBeatName": "Complication",
      "parentBeatSlug": "complication"
    }
  ],
  "loops": [
    {
      "to": 44,
      "from": 40,
      "name": "Pattern Hunter",
      "slug": "02-pattern-hunter",
      "bestFor": "Time-pressed audiences, building consensus, when data is strong",
      "matchId": "019dd95a-07fe-70ce-8d3e-3752140f3c2d",
      "evidence": "VLMs, diffusion image, Stable Video Diffusion, gated text-to-video, Movie Gen audio",
      "position": 6,
      "objective": "Map progress in multimodal vision/video/audio generation",
      "structure": "Evidence A -> Evidence B -> Evidence C -> Pattern/Conclusion",
      "confidence": 76,
      "description": "Group multiple pieces of evidence that together point to a pattern or conclusion"
    }
  ],
  "imagePathAlt": null,
  "thumbSrc": null,
  "thumbSrcAlt": null,
  "locked": true
}