{
  "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": 66,
  "pageCount": 313,
  "prevPage": 65,
  "nextPage": 67,
  "slideType": "case_study",
  "function": "illustrate_case",
  "density": "dense",
  "nDataPoints": 8,
  "notes": "The slide presents performance metrics for AMIE vs PCPs using line charts showing accuracy over top-k.",
  "elementsJson": [
    "headline_text",
    "bullet_list",
    "line_chart"
  ],
  "metadataConfidence": 1,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5d-0d0f98caffe1/66",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5d-0d0f98caffe1",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5d-0d0f98caffe1.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5d-0d0f98caffe1#slide-66",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": "A specialized clinical dialogue model beats unassisted doctors on NEJM-grade diagnosis, outperforms primary care physicians (PCPs) in (multimodal) simulated consults, is non-inferior on multi-visit disease management, and outperforms PCPs in history taking and medical note writing under oversight.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-41ac-7180-a821-5f5f499d7fbd",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.55,
        "w": 0.25,
        "x": 0.72,
        "y": 0.34
      },
      "kind": "chart",
      "text": "Performance charts showing accuracy vs top-k for AMIE and PCPs",
      "attrs": null,
      "subkind": "line",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "d9fecf8e-d4af-42df-b5c3-aa6d02fc1be4",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.55,
        "w": 0.65,
        "x": 0.05,
        "y": 0.36
      },
      "kind": "list",
      "text": "AMIE is a multimodal clinical dialogue model trained with simulated self-play consultations, equipped with inference-time chain-of-thought, retrieval of guidelines, and a custom-built clinician cockpit for oversight.\nOn 302 real-world NEJM cases, AMIE hits 59.1% top-10 vs 33.6% for unassisted clinicians, 44.5% assisted by search, 51.8% assisted by AMIE.\nIn randomized, double-blind OSCE-style consults that assess clinical competence, physicians and patient actors rated AMIE above PCPs on the majority of evaluation axes, incl. higher diagnostic accuracy (159 scenarios).\nThis includes better management reasoning across multiple visits (100 scenarios), better use of multimodal artifacts and (105 scenarios) and better history taking, medical notes and composite performance in an AMIE+clinician team (60 scenarios).",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "ef72f6ff-5d62-47ee-abc8-d8ab64a1ba92",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": null,
      "kind": "metric",
      "text": "Diagnostic Accuracy (%): 59.1%",
      "attrs": null,
      "subkind": "primary",
      "toolName": "Quantification",
      "toolSlug": "quantification",
      "confidence": null,
      "componentId": "019dd952-41ac-7180-a821-6009bb9f5b3c",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.12,
        "w": 0.65,
        "x": 0.05,
        "y": 0.22
      },
      "kind": "paragraph",
      "text": "A specialized clinical dialogue model beats unassisted doctors on NEJM-grade diagnosis, outperforms primary care physicians (PCPs) in (multimodal) simulated consults, is non-inferior on multi-visit disease management, and outperforms PCPs in history taking and medical note writing under oversight.",
      "attrs": null,
      "subkind": "paragraph",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "222170e6-4562-4eb2-8cc6-2d9769b72a6f",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.05,
        "w": 0.9,
        "x": 0.05,
        "y": 0.14
      },
      "kind": "title",
      "text": "AMIE address multimodal diagnostic consultations in longitudinal care and w/oversight",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "667909f1-a006-4847-b375-7c173b6c59ee",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [
    {
      "name": "Monroe's Motivated Sequence",
      "slug": "monroes-motivated-sequence",
      "agent": "Storyteller",
      "layer": "slide",
      "matchId": "f529a56e-0f48-40dc-93bb-827fc1ce2e7e",
      "evidence": "The slide mentions a multimodal clinical dialogue model trained with simulated self-play consultations.",
      "confidence": 0.7
    }
  ],
  "frameworks": [],
  "arcBeats": [
    {
      "to": 189,
      "from": 12,
      "beatId": "019dd95a-0682-776c-8e35-9f380673831f",
      "arcName": "The Triple Take",
      "arcSlug": "triple-take",
      "beatName": "The Facts (What)",
      "beatSlug": "triple-take-the-facts-what",
      "evidence": "Sections 1-2 lay out research findings and industry data with charts and case studies.",
      "position": 1,
      "confidence": 70,
      "parentBeatName": "Setup",
      "parentBeatSlug": "setup"
    },
    {
      "to": 89,
      "from": 12,
      "beatId": "019dd95a-0682-776c-8e35-ac68557d958b",
      "arcName": "The Onion",
      "arcSlug": "onion",
      "beatName": "First Layer",
      "beatSlug": "onion-first-layer",
      "evidence": "Research section unpacks technical layer beneath the headlines.",
      "position": 2,
      "confidence": 45,
      "parentBeatName": "Development",
      "parentBeatSlug": "development"
    }
  ],
  "loops": [
    {
      "to": 67,
      "from": 55,
      "name": "Zoom In",
      "slug": "06-zoom-in",
      "bestFor": "Technical deep-dives, case studies, detailed analysis",
      "matchId": "019dd95a-07fe-70ce-8d3e-9d32ca243ef1",
      "evidence": "Sequence of case studies: AlphaEvolve, UMA, MatterGen, AMIE, AlphaFold reproductions all illustrate AI-as-scientist.",
      "position": 5,
      "objective": "Survey AI's rise as scientific collaborator across biology/chemistry/medicine",
      "structure": "The Big Picture -> Key Area of Focus -> Specific Detail -> Implication",
      "confidence": 75,
      "description": "Start broad, then progressively focus on specific details that prove your point"
    }
  ],
  "imagePathAlt": null,
  "thumbSrc": null,
  "thumbSrcAlt": null,
  "locked": true
}