{
  "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": 258,
  "pageCount": 313,
  "prevPage": 257,
  "nextPage": 259,
  "slideType": "industry_trends",
  "function": "diagnose",
  "density": "overcrowded",
  "nDataPoints": 9,
  "notes": "The chart shows 'Delusion confirmation rating' across various AI models.",
  "elementsJson": [
    "headline_text",
    "bullet_list",
    "bar_chart_horizontal"
  ],
  "metadataConfidence": 1,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5d-0d0f98caffe1/258",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5d-0d0f98caffe1",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5d-0d0f98caffe1.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5d-0d0f98caffe1#slide-258",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": "Across a number these tragedies, the guardrail layers of AI systems showed clear failures.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-47c3-7407-8d07-99623dbed209",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.35,
        "w": 0.38,
        "x": 0.58,
        "y": 0.485
      },
      "kind": "chart",
      "text": "Many AIs Encourage Users' Delusions",
      "attrs": null,
      "subkind": "bar-horizontal",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "cc85e9fa-7881-4fb5-b205-fa88182601d5",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.15,
        "w": 0.45,
        "x": 0.02,
        "y": 0.63
      },
      "kind": "list",
      "text": "Are these isolated incidents or are chatbots causing a widespread crisis? Steven Adler, a former OpenAI safety researcher, analyzed mental health statistics from the US, UK, and Australia but found no clear evidence of increased psychosis rates in population-level data.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "0d05c0a5-1db6-475a-8798-86df59c077e8",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.15,
        "w": 0.45,
        "x": 0.02,
        "y": 0.47
      },
      "kind": "list",
      "text": "Labs face exposure to new liabilities as legal battles unfold due to AI-assisted suicides. This has prompted new controls (e.g. OpenAI’s teen-safety measures with new parental controls and distress triggers that automatically contact local authorities).",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "5a33fc42-a90e-4bf7-af69-913e1031b811",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.08,
        "w": 0.45,
        "x": 0.02,
        "y": 0.21
      },
      "kind": "list",
      "text": "High-profile cases of AI psychosis, instances where AI interactions worsen or induce adverse psychological symptoms, continue to rise across the globe.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "6460862e-0903-4146-ac41-9a5ba7a389c2",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.15,
        "w": 0.45,
        "x": 0.02,
        "y": 0.31
      },
      "kind": "list",
      "text": "Across a number these tragedies, the guardrail layers of AI systems showed clear failures. Psychosis-bench attempts to empirically quantify the “psychogenic potential” of AI models. But results find current AI systems to display overt sycophancy and inadequate crisis support, which can reinforce users’ delusional beliefs.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "b91f516f-f8a2-412f-98ec-6846a5578513",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": null,
      "kind": "metric",
      "text": "Delusion confirmation rating",
      "attrs": null,
      "subkind": "primary",
      "toolName": "Quantification",
      "toolSlug": "quantification",
      "confidence": null,
      "componentId": "019dd952-47c3-7407-8d07-9d8d9a5f947b",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.04,
        "w": 0.45,
        "x": 0.02,
        "y": 0.14
      },
      "kind": "title",
      "text": "The concerning phenomenon of AI psychosis",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "9e5082ef-ef24-4264-a039-e8fe9683b8d9",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [
    {
      "name": "Emotional Appeal",
      "slug": "emotional-appeal",
      "agent": "Storyteller",
      "layer": "slide",
      "matchId": "05c4a5d6-4d79-4179-be49-7cdf3bc44731",
      "evidence": "High-profile cases of AI psychosis, instances where AI interactions worsen or induce adverse psychological symptoms, continue to rise across the globe.",
      "confidence": 0.6
    }
  ],
  "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
}