{
  "docId": "019dd923-5e88-73ef-bd5d-06b04d219fea",
  "docSlug": "dd91c78f6570bf29",
  "documentTitle": "2023 Air Street Capital The State of AI Report 2023",
  "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": 105,
  "pageCount": 163,
  "prevPage": 104,
  "nextPage": 106,
  "slideType": "case_study",
  "function": "illustrate_case",
  "density": "dense",
  "nDataPoints": 2,
  "notes": "Includes a list of authors and logos of companies they founded or joined.",
  "elementsJson": [
    "paragraph",
    "line_chart",
    "logo_grid"
  ],
  "metadataConfidence": 1,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5d-06b04d219fea/105",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5d-06b04d219fea",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5d-06b04d219fea.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5d-06b04d219fea#slide-105",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": "A 2017 paper from the same lab, \"Deep learning scaling is predictable, empirically\" demonstrated early evidence for \"scaling laws\", which now underpins the large-scale AI we see and use today.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-47c4-7719-a22b-176c30163b95",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.25,
        "w": 0.4,
        "x": 0.02,
        "y": 0.65
      },
      "kind": "chart",
      "text": "Learning curve and model size results and trends for word language models.",
      "attrs": null,
      "subkind": "line",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "c43d6d34-5f9f-46de-915f-b21927a40cd5",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.3,
        "w": 0.15,
        "x": 0.82,
        "y": 0.6
      },
      "kind": "image",
      "text": "Logos of Meta, Google, Reka, Lamini, NVIDIA, Adept AI",
      "attrs": null,
      "subkind": "logo-grid",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "08d40416-3c85-4dde-ab83-c6c267cd85fc",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.15,
        "w": 0.4,
        "x": 0.42,
        "y": 0.82
      },
      "kind": "list",
      "text": "Dario Amodei, Rishita Anubhai, Eric Battenberg, Carl Case, Jared Casper, Bryan Catanzaro, Jingdong Chen, Mike Chrzanowski, Adam Coates, Greg Diamos, Erich Elsen, Jesse Engel, Linxi Fan, Christopher Fougner, Tony Han, Awni Hannun, Billy Jun, Patrick LeGresley, Libby Lin, Sharan Narang, Andrew Ng, Sherjil Ozair, Ryan Prenger, Jonathan Raiman, Sanjeev Satheesh, David Seetapun, Shubho Sengupta, Yi Wang, Zhiqian Wang, Chong Wang, Bo Xiao, Dani Yogatama, Jun Zhan, Zhenyao Zhu",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "c6dbb0ef-18b4-4e8e-9945-d33474a25232",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.35,
        "w": 0.96,
        "x": 0.02,
        "y": 0.21
      },
      "kind": "paragraph",
      "text": "In 2015, Baidu's Silicon Valley AI Lab introduced a fully end-to-end deep learning based system for speech recognition. The work did away with hand-crafted feature-based pipelines and heavy use of computation: “Key to our approach is our application of HPC techniques, resulting in a 7x speedup over our previous system [...] Our system is competitive with the transcription of human workers when benchmarked on standard datasets.” A 2017 paper from the same lab, “Deep learning scaling is predictable, empirically” demonstrated early evidence for “scaling laws”, which now underpins the large-scale AI we see and use today. Many DS2 authors have gone onto be founders or execs of leading ML companies, often leading their large scale efforts in language modeling and related fields.",
      "attrs": null,
      "subkind": "paragraph",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "02a87bcd-2dc7-494e-ac56-198ac5dfbc6f",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": null,
      "kind": "quote",
      "text": "Key to our approach is our application of HPC techniques, resulting in a 7x speedup over our previous system [...] Our system is competitive with the transcription of human workers when benchmarked on standard datasets.",
      "attrs": null,
      "subkind": null,
      "toolName": "Authority citation",
      "toolSlug": "authority-citation",
      "confidence": null,
      "componentId": "019dd952-47c4-7719-a22b-18e3acdbc806",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.04,
        "w": 0.45,
        "x": 0.02,
        "y": 0.14
      },
      "kind": "title",
      "text": "DeepSpeech 2: The early masters of scale",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "b48b6ced-e4bf-4de2-829d-b19edd5035d7",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [],
  "frameworks": [],
  "arcBeats": [
    {
      "to": 120,
      "from": 11,
      "beatId": "019dd95a-0682-776c-8e35-41afd44ef59f",
      "arcName": "The Triple Take",
      "arcSlug": "triple-take",
      "beatName": "The Facts (What)",
      "beatSlug": "triple-take-the-facts-what",
      "evidence": "Research + Industry sections inventory model, compute, funding facts.",
      "position": 1,
      "confidence": 78,
      "parentBeatName": "Setup",
      "parentBeatSlug": "setup"
    },
    {
      "to": 120,
      "from": 11,
      "beatId": "019dd95a-0682-776c-8e35-523bfb7f96e6",
      "arcName": "The Mountain",
      "arcSlug": "mountain",
      "beatName": "Rising Action",
      "beatSlug": "mountain-rising-action",
      "evidence": "Escalating capabilities, compute concentration and capital flows.",
      "position": 2,
      "confidence": 45,
      "parentBeatName": "Development",
      "parentBeatSlug": "development"
    }
  ],
  "loops": [
    {
      "to": 120,
      "from": 99,
      "name": "Pattern Hunter",
      "slug": "02-pattern-hunter",
      "bestFor": "Time-pressed audiences, building consensus, when data is strong",
      "matchId": "019dd95a-07fe-70ce-8d3e-09a2221c6f1a",
      "evidence": "Funding, valuations, geo split, unicorns, sector mix, M&A, CVC, GenAI rounds compound the same thesis.",
      "position": 11,
      "objective": "Map the GenAI investment surge across stages, geos and corporates",
      "structure": "Evidence A -> Evidence B -> Evidence C -> Pattern/Conclusion",
      "confidence": 80,
      "description": "Group multiple pieces of evidence that together point to a pattern or conclusion"
    }
  ],
  "imagePathAlt": null,
  "thumbSrc": null,
  "thumbSrcAlt": null,
  "locked": true
}