{
  "docId": "019dd923-5e88-73ef-bd5d-0d0f98caffe1",
  "docSlug": "5df6bd1b0447b5f6",
  "documentTitle": "2025 Air Street Capital The State of AI Report 2025",
  "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": 248,
  "pageCount": 313,
  "prevPage": 247,
  "nextPage": 249,
  "slideType": "comparison_table",
  "function": "quantify_impact",
  "density": "overcrowded",
  "nDataPoints": 2,
  "notes": "The chart is a simplified bar comparison showing the scale difference between $92B (AI Labs) and $133M (External Testing).",
  "elementsJson": [
    "headline_text",
    "bullet_list",
    "bar_chart_horizontal",
    "footnote"
  ],
  "metadataConfidence": 1,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5d-0d0f98caffe1/248",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5d-0d0f98caffe1",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5d-0d0f98caffe1.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5d-0d0f98caffe1#slide-248",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": "This creates a structural conflict of interest: findings that call for caution may be deprioritized in favor of speed and market advantage.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-47c3-7407-8d06-5e8300f48473",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.3,
        "w": 0.4,
        "x": 0.55,
        "y": 0.5
      },
      "kind": "chart",
      "text": "Spending Gap Between AI Labs and External Testers in America",
      "attrs": null,
      "subkind": "bar-horizontal",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "99956c26-9d5d-4842-9c4f-781ff65b1f4e",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.15,
        "w": 0.45,
        "x": 0.05,
        "y": 0.65
      },
      "kind": "list",
      "text": "This isn't (just) about money: external orgs also lack other means to attract talent like comparable prestige, and access to privileged information / pre-release models. As such it is difficult for them to provide a credible counterweight, leaving the ecosystem over-reliant on self-policing.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "27792c15-c1f5-4ba1-9c91-25dd4e0ab5f0",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.15,
        "w": 0.45,
        "x": 0.05,
        "y": 0.48
      },
      "kind": "list",
      "text": "Although well-resourced, internal safety teams ultimately answer to the same organizations racing to commercialize frontier models. This creates a structural conflict of interest: findings that call for caution may be deprioritized in favor of speed and market advantage.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "3e5c5bf5-975f-4448-809f-6e91070cc622",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.15,
        "w": 0.45,
        "x": 0.05,
        "y": 0.33
      },
      "kind": "list",
      "text": "We estimate the eleven most prominent American AI safety-science organizations combined will spend just $133.4M in 2025. This grouping includes the following organizations: CAISI, METR, CAIS, FAR.AI, Haize Labs, Palisade Research, Virtue AI, Gray Swan, Redwood Research, Irregular, and the Frontier Model Forum.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "95fc9191-d328-4bd7-98cb-9b8221198133",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": null,
      "kind": "metric",
      "text": "Estimated 2025 Spend: $92B",
      "attrs": null,
      "subkind": "primary",
      "toolName": "Quantification",
      "toolSlug": "quantification",
      "confidence": null,
      "componentId": "019dd952-47c3-7407-8d06-60d01d314b48",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.05,
        "w": 0.4,
        "x": 0.55,
        "y": 0.83
      },
      "kind": "source-note",
      "text": "* 'AI Labs' corresponds to a rough estimate of each lab's total expenditures in 2025 (compute, personnel costs, other opex)",
      "attrs": null,
      "subkind": null,
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "ab6a3fac-d301-427a-8cb8-704e11520e8d",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.05,
        "w": 0.7,
        "x": 0.05,
        "y": 0.14
      },
      "kind": "title",
      "text": "AI labs spend more in a day than AI safety science organizations spend in a year",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "e81edbd2-f2e9-4186-aa32-c80cff603a6d",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [
    {
      "name": "Chart Selection Guide",
      "slug": "chart-selection-guide",
      "agent": "Designer",
      "layer": "slide",
      "matchId": "702dedd3-2be9-4cdd-b5ab-949bc27b09fb",
      "evidence": "The slide includes a bar chart showing the spending gap between AI labs and external testers.",
      "confidence": 0.6
    },
    {
      "name": "Comparison frame",
      "slug": "comparison-frame",
      "agent": null,
      "layer": "slide",
      "matchId": "ffc84980-a960-4039-bae7-92e2c43c73f5",
      "evidence": "The slide compares the spending of AI labs and external safety testers.",
      "confidence": 0.7
    },
    {
      "name": "Concrete Language",
      "slug": "concrete-language",
      "agent": "Storyteller",
      "layer": "slide",
      "matchId": "019dd95a-1055-74e0-a95c-c3457132076a",
      "evidence": "'AI labs spend more in a day than safety orgs in a year' — concrete asymmetry.",
      "confidence": 85
    },
    {
      "name": "Emotional Appeal",
      "slug": "emotional-appeal",
      "agent": "Storyteller",
      "layer": "slide",
      "matchId": "019dd95a-1055-74e0-a95c-c420c91ed8ce",
      "evidence": "Frames 'structural conflict of interest' to provoke concern.",
      "confidence": 65
    }
  ],
  "frameworks": [],
  "arcBeats": [
    {
      "to": 282,
      "from": 190,
      "beatId": "019dd95a-0682-776c-8e35-a256fe16ea9c",
      "arcName": "The Triple Take",
      "arcSlug": "triple-take",
      "beatName": "The Implications (So What)",
      "beatSlug": "triple-take-the-implications-so-what",
      "evidence": "Sections 3-4 frame political and safety consequences of the technical/industry trends.",
      "position": 2,
      "confidence": 70,
      "parentBeatName": "Reflection",
      "parentBeatSlug": "reflection"
    },
    {
      "to": 282,
      "from": 190,
      "beatId": "019dd95a-0682-776c-8e35-b713af60bf88",
      "arcName": "The Onion",
      "arcSlug": "onion",
      "beatName": "Core Insight",
      "beatSlug": "onion-core-insight",
      "evidence": "Politics and Safety reveal the structural risks at the core.",
      "position": 4,
      "confidence": 45,
      "parentBeatName": "Turn",
      "parentBeatSlug": "turn"
    }
  ],
  "loops": [
    {
      "to": 282,
      "from": 247,
      "name": "Pre Mortem",
      "slug": "19-pre-mortem",
      "bestFor": "Project kick-offs, risk management, building trust with skeptical stakeholders",
      "matchId": "019dd95a-07fe-70ce-8d3e-bb4d43115ab4",
      "evidence": "Vibe hacking, AI psychosis, alignment faking, scheming, prompt injection — what could go wrong + mitigations.",
      "position": 12,
      "objective": "Surface AI safety failure modes before they become catastrophic",
      "structure": "The Future Disaster (Hypothetical) -> What Went Wrong -> The Prevention Plan",
      "confidence": 75,
      "description": "Build trust by anticipating failure modes before they happen and solving them in advance"
    }
  ],
  "imagePathAlt": null,
  "thumbSrc": null,
  "thumbSrcAlt": null,
  "locked": true
}