{
  "docId": "019dd923-5e88-73ef-bd5c-fd5012384ae3",
  "docSlug": "bf350dd574c19997",
  "documentTitle": "2021 Air Street Capital The State of AI Report 2021",
  "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": 164,
  "pageCount": 188,
  "prevPage": 163,
  "nextPage": 165,
  "slideType": "case_study",
  "function": "illustrate_case",
  "density": "dense",
  "nDataPoints": 3,
  "notes": "The slide uses a treemap to visualize the composition of 'The Pile' dataset.",
  "elementsJson": [
    "headline_text",
    "bullet_list",
    "tree_map"
  ],
  "metadataConfidence": 1,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5c-fd5012384ae3/164",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5c-fd5012384ae3",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5c-fd5012384ae3.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5c-fd5012384ae3#slide-164",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": "A notable achievement of the project has been to create The Pile, a free and publicly released 800GB dataset of diverse English text for large language modelling.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-47c4-7719-a211-afef95c18392",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.45,
        "w": 0.35,
        "x": 0.62,
        "y": 0.35
      },
      "kind": "chart",
      "text": "Composition of the Pile by Category",
      "attrs": null,
      "subkind": "treemap",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "c4f2edce-0036-495c-9c34-42930c2518cd",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.55,
        "w": 0.55,
        "x": 0.03,
        "y": 0.32
      },
      "kind": "list",
      "text": "Eleuther’s noble aims have attracted support from the wider community attracting 10,000 members to their Discord server.\nCoreWeave, a cloud service provider that specialises in high performance ML, contributed compute to train the GPT-Neo models. The EleutherAI team created a way to split AI computations across multiple machines.\nThe collective followed this with GPT-J-6B, a 6B parameter model for use with a new codebase, Mesh Transformer JAX.\nThe EleutherAI language models hosted on Hugging Face had over 500,000 downloads in August 2021.\nThe EleutherAI community has now expanded its activity and is working on open-source alternatives in BioML and generative art.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "40bc9328-071a-467e-853e-ffe94925bd38",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": null,
      "kind": "metric",
      "text": "Dataset size: 800GB",
      "attrs": null,
      "subkind": "primary",
      "toolName": "Quantification",
      "toolSlug": "quantification",
      "confidence": null,
      "componentId": "019dd952-47c4-7719-a211-b02474f9dd01",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.08,
        "w": 0.55,
        "x": 0.03,
        "y": 0.22
      },
      "kind": "paragraph",
      "text": "A notable achievement of the project has been to create The Pile, a free and publicly released 800GB dataset of diverse English text for large language modelling.",
      "attrs": null,
      "subkind": "paragraph",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "7334823c-30fc-4b0a-b183-983851a086cb",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.05,
        "w": 0.7,
        "x": 0.03,
        "y": 0.14
      },
      "kind": "title",
      "text": "AI Governance: EleutherAI mounts attempt to decentralise power via open source",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "cfd7ec65-55ee-47ed-addb-ab8c83bd53d6",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [],
  "frameworks": [],
  "arcBeats": [
    {
      "to": 181,
      "from": 153,
      "beatId": "019dd95a-0682-776c-8e35-0cadb21b77bf",
      "arcName": "The Triple Take",
      "arcSlug": "triple-take",
      "beatName": "The Implications (So What)",
      "beatSlug": "triple-take-the-implications-so-what",
      "evidence": "Section 4 Politics: AI Safety, Governance, Regulation, Military implications.",
      "position": 2,
      "confidence": 78,
      "parentBeatName": "Reflection",
      "parentBeatSlug": "reflection"
    },
    {
      "to": 181,
      "from": 153,
      "beatId": "019dd95a-0682-776c-8e35-1e37fb3155df",
      "arcName": "The Mountain",
      "arcSlug": "mountain",
      "beatName": "Climax",
      "beatSlug": "mountain-climax",
      "evidence": "Politics section: ethics firings, AI Act, military AI deployments.",
      "position": 3,
      "confidence": 55,
      "parentBeatName": "Turn",
      "parentBeatSlug": "turn"
    }
  ],
  "loops": [
    {
      "to": 165,
      "from": 160,
      "name": "Pattern Hunter",
      "slug": "02-pattern-hunter",
      "bestFor": "Time-pressed audiences, building consensus, when data is strong",
      "matchId": "019dd95a-07fe-70ce-8d3d-88b20d80074d",
      "evidence": "DeepMind independence fail, Anthropic, EleutherAI open source, BigScience hybrid model.",
      "position": 19,
      "objective": "Survey AI governance experiments",
      "structure": "Evidence A -> Evidence B -> Evidence C -> Pattern/Conclusion",
      "confidence": 72,
      "description": "Group multiple pieces of evidence that together point to a pattern or conclusion"
    }
  ],
  "imagePathAlt": null,
  "thumbSrc": null,
  "thumbSrcAlt": null,
  "locked": true
}