{
  "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": 75,
  "pageCount": 213,
  "prevPage": 74,
  "nextPage": 76,
  "slideType": "industry_trends",
  "function": "summarize",
  "density": "dense",
  "nDataPoints": 0,
  "notes": "The slide uses a narrative structure to highlight four specific technical developments in robotics by Google DeepMind.",
  "elementsJson": [
    "headline_text",
    "bullet_list",
    "photo"
  ],
  "metadataConfidence": 1,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5d-0856e1444fb9/75",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5d-0856e1444fb9",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5d-0856e1444fb9.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5d-0856e1444fb9#slide-75",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": "Despite all eyes being on Gemini, the Google DeepMind team has steadily been increasing its robotics output, improving the efficiency, adaptability, and data collection of robots.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-47c3-7407-8d15-59e7e431ac7f",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.35,
        "w": 0.32,
        "x": 0.65,
        "y": 0.48
      },
      "kind": "image",
      "text": "Photo of a robot in an office environment with a person sitting at a desk.",
      "attrs": null,
      "subkind": "photo",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "d561f0a1-1d8d-4031-9036-e3222c47ce39",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.55,
        "w": 0.6,
        "x": 0.02,
        "y": 0.32
      },
      "kind": "list",
      "text": "The team created AutoRT, a system that uses a VLM for environmental understanding and an LLM to suggest a list of creative tasks the robot could carry out. These models are then combined with a robot control policy. This helps to scale up deployment quickly in previously unseen environments.\nRT-Trajectory enhances robotic learning through video input. For each video in the dataset of demonstrations, a 2D sketch of the gripper performing the task is overlaid. This provide practical visual hits to the model as it learns.\nThe team have also improved the efficiency of transformers. SARA-RT is a novel 'up-training' method to convert pre-trained or fine-tuned robotic policies from quadratic to linear attention, while maintaining quality.\nResearchers have found Gemini 1.5 Pro's multimodal capabilities and long context window makes it an effective way of interacting with robots via natural language.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "5f9cc794-6af5-4daf-ae34-a4bd5c63ae98",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.08,
        "w": 0.6,
        "x": 0.02,
        "y": 0.22
      },
      "kind": "paragraph",
      "text": "Despite all eyes being on Gemini, the Google DeepMind team has steadily been increasing its robotics output, improving the efficiency, adaptability, and data collection of robots.",
      "attrs": null,
      "subkind": "paragraph",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "dc8040c2-ecd2-4a8a-bb35-a1800cd9c103",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.05,
        "w": 0.5,
        "x": 0.02,
        "y": 0.14
      },
      "kind": "title",
      "text": "Google DeepMind quietly emerges as a robotics leader",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "efe833c5-943d-48a5-a1e4-cd71083cdb42",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [],
  "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": 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": 81,
      "from": 72,
      "name": "Pattern Hunter",
      "slug": "02-pattern-hunter",
      "bestFor": "Time-pressed audiences, building consensus, when data is strong",
      "matchId": "019dd95a-07fe-70ce-8d3e-431790a1bfd7",
      "evidence": "Wayve LINGO-2, SAM2, DeepMind robotics, LeRobot, diffusion policies, humanoids, Vision Pro",
      "position": 9,
      "objective": "Show robotics renaissance via converging evidence",
      "structure": "Evidence A -> Evidence B -> Evidence C -> Pattern/Conclusion",
      "confidence": 78,
      "description": "Group multiple pieces of evidence that together point to a pattern or conclusion"
    }
  ],
  "imagePathAlt": null,
  "thumbSrc": null,
  "thumbSrcAlt": null,
  "locked": true
}