{
  "docId": "019dd923-5e88-73ef-bd5c-f6fdd8f57895",
  "docSlug": "3a892393f72c1ff5",
  "documentTitle": "2019 Air Street Capital The State of AI Report 2019",
  "authorId": "AirStreetCapital",
  "authorName": "Nathan Benaich and Ian Hogarth",
  "documentKindSlug": "consulting-deck",
  "documentKindLabel": "Consulting deck",
  "sourceTypeSlug": "vc_research",
  "sourceTypeLabel": "VC research",
  "presentationDate": null,
  "orientation": "landscape",
  "aspectRatio": 1.777,
  "pageNumber": 23,
  "pageCount": 136,
  "prevPage": 22,
  "nextPage": 24,
  "slideType": "industry_trends",
  "function": "summarize",
  "density": "dense",
  "nDataPoints": 5,
  "notes": "The slide tracks the evolution of NLP model performance against the GLUE benchmark, noting the introduction of SUPERglue due to rapid progress.",
  "elementsJson": [
    "headline_text",
    "bullet_list",
    "bar_chart_vertical"
  ],
  "metadataConfidence": 1,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5c-f6fdd8f57895/23",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5c-f6fdd8f57895",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5c-f6fdd8f57895.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5c-f6fdd8f57895#slide-23",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": "As a demonstration of how quickly progress is being made in NLP, the state-of-the art has increased from a score of 69 to 88 over 13 months.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-41ab-72d9-bbc3-1a6375ff3c06",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.55,
        "w": 0.3,
        "x": 0.65,
        "y": 0.25
      },
      "kind": "chart",
      "text": "GLUE benchmark progress chart showing scores of 69, 73, 81, 88 against a human baseline of 87.",
      "attrs": null,
      "subkind": "bar-vertical",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "1cf61668-a737-405f-8595-57d88df5795d",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.5,
        "w": 0.6,
        "x": 0.05,
        "y": 0.28
      },
      "kind": "list",
      "text": "Human understanding of language is general and flexible. The GLUE benchmark provides a single benchmark for evaluating NLP systems at a range of tasks spanning logic, common sense understanding, and lexical semantics. The right hand charts progress on the leaderboard.\nThe benchmark is designed to favor systems that share general linguistic knowledge across tasks.\nAs a demonstration of how quickly progress is being made in NLP, the state-of-the-art has increased from a score of 69 to 88 over 13 months. The human baseline level is 87.\nProgress was so much faster than anticipated that a new benchmark SUPERglue has already been introduced.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "f14e9377-f045-44bd-a34a-779057644dc3",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": null,
      "kind": "metric",
      "text": "GLUE benchmark score: 88",
      "attrs": null,
      "subkind": "primary",
      "toolName": "Quantification",
      "toolSlug": "quantification",
      "confidence": null,
      "componentId": "019dd952-41ab-72d9-bbc3-1e04516c03a8",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.04,
        "w": 0.6,
        "x": 0.05,
        "y": 0.22
      },
      "kind": "paragraph",
      "text": "New performance benchmark GLUE tests performance at a variety of language understanding tasks.",
      "attrs": null,
      "subkind": "paragraph",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "9e1849f6-1a2c-4cad-ab5f-a6472f376579",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.05,
        "w": 0.6,
        "x": 0.05,
        "y": 0.14
      },
      "kind": "title",
      "text": "A big year in natural language processing: Researchers start sniffing GLUE",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "dec8eb6e-d179-4660-af3d-53b3f3ddaece",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [
    {
      "name": "Action Titles",
      "slug": "action-titles",
      "agent": "Architect",
      "layer": "slide",
      "matchId": "019dd95a-0fd5-7148-8ecc-d20aab368ff9",
      "evidence": "Title quantifies GLUE jump 69→88 in 13 months.",
      "confidence": 80
    },
    {
      "name": "Chart Selection Guide",
      "slug": "chart-selection-guide",
      "agent": "Designer",
      "layer": "slide",
      "matchId": "019dd95a-0fd5-7148-8ecc-d56235a04011",
      "evidence": "Score-progression chart over time.",
      "confidence": 60
    },
    {
      "name": "Chart Selection Guide",
      "slug": "chart-selection-guide",
      "agent": "Designer",
      "layer": "slide",
      "matchId": "baf88412-e13d-4fb5-9627-42462391743f",
      "evidence": "GLUE benchmark progress chart showing scores of 69, 73, 81, 88 against a human baseline of 87.",
      "confidence": 0.6
    }
  ],
  "frameworks": [],
  "arcBeats": [
    {
      "to": 92,
      "from": 5,
      "beatId": "019dd95a-0682-776c-8e34-ce57e5bbe574",
      "arcName": "The Triple Take",
      "arcSlug": "triple-take",
      "beatName": "The Facts (What)",
      "beatSlug": "triple-take-the-facts-what",
      "evidence": "Definitions through Research/Talent/Industry — fact-dense case studies and benchmarks.",
      "position": 1,
      "confidence": 70,
      "parentBeatName": "Setup",
      "parentBeatSlug": "setup"
    },
    {
      "to": 92,
      "from": 9,
      "beatId": "019dd95a-0682-776c-8e34-ddbccb391af1",
      "arcName": "Voyage and Return",
      "arcSlug": "voyage-return",
      "beatName": "The Unknown",
      "beatSlug": "voyage-return-the-unknown",
      "evidence": "Research/Talent/Industry frontiers explored.",
      "position": 2,
      "confidence": 45,
      "parentBeatName": "Complication",
      "parentBeatSlug": "complication"
    }
  ],
  "loops": [
    {
      "to": 25,
      "from": 21,
      "name": "Pattern Hunter",
      "slug": "02-pattern-hunter",
      "bestFor": "Time-pressed audiences, building consensus, when data is strong",
      "matchId": "019dd95a-07fe-70ce-8d3c-a1ab93896461",
      "evidence": "Five contiguous slides titled 'A big year in natural language processing'.",
      "position": 4,
      "objective": "Establish 2019 as NLP's breakthrough year via pretrained-LM evidence",
      "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
}