{
  "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": 135,
  "pageCount": 213,
  "prevPage": 134,
  "nextPage": 136,
  "slideType": "industry_trends",
  "function": "summarize",
  "density": "overcrowded",
  "nDataPoints": 5,
  "notes": "Includes a screenshot of the PolyAI website as a visual example.",
  "elementsJson": [
    "headline_text",
    "bullet_list",
    "screenshot"
  ],
  "metadataConfidence": 1,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5d-0856e1444fb9/135",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5d-0856e1444fb9",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5d-0856e1444fb9.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5d-0856e1444fb9#slide-135",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": "While text-to-speech benefits from a 'wow effect', speech recognition has the potential to automate away mundane tasks at scale.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-47c3-7407-8d1b-500043778e75",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.285,
        "w": 0.255,
        "x": 0.695,
        "y": 0.49
      },
      "kind": "image",
      "text": "PolyAI website screenshot",
      "attrs": null,
      "subkind": "screenshot",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "3a4d9a18-4b4d-44e1-bd02-a5d7380a26fb",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.15,
        "w": 0.65,
        "x": 0.03,
        "y": 0.47
      },
      "kind": "list",
      "text": "These start-ups are focused on plugging shortages of call center staff and allowing for more natural speech from customers, including corrections, hesitation, interruption, and topic changes - areas traditional automated systems have struggled with.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "29110dfe-3b83-4eb9-a9f9-67545181ea32",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.12,
        "w": 0.65,
        "x": 0.03,
        "y": 0.72
      },
      "kind": "list",
      "text": "For example, Assembly AI has built Universal-1, a multilingual model trained on 12.5M hours of speech that runs faster, with less compute, fewer errors and better ambient noise reduction than OpenAI's Whisper.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "2d987dcd-f15a-4992-93f1-78b7c6d655f0",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.15,
        "w": 0.65,
        "x": 0.03,
        "y": 0.31
      },
      "kind": "list",
      "text": "A string of start-ups working to use speech recognition for a range of use cases including customer support and call centers have scored funding rounds in the last year or so, including Assembly AI ($50M), Deepgram ($72M), PolyAI ($50M), Parloa ($66M). PolyAI's revenue is set to triple this year.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "bad5a7e2-88ea-418b-82a7-db7c6c2c8afb",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.07,
        "w": 0.65,
        "x": 0.03,
        "y": 0.64
      },
      "kind": "list",
      "text": "While AI-powered transcription and audio analysis isn't new, accuracy is improving as a result of larger datasets, transformer models.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "c571010d-7f8f-4d52-9bbb-31336fa9bc18",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": null,
      "kind": "metric",
      "text": "funding: $72M",
      "attrs": null,
      "subkind": "primary",
      "toolName": "Quantification",
      "toolSlug": "quantification",
      "confidence": null,
      "componentId": "019dd952-47c3-7407-8d1b-57103d81a509",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.08,
        "w": 0.94,
        "x": 0.03,
        "y": 0.21
      },
      "kind": "paragraph",
      "text": "While text-to-speech benefits from a 'wow effect', speech recognition has the potential to automate away mundane tasks at scale. Investors are beginning to see its potential to scale.",
      "attrs": null,
      "subkind": "paragraph",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "38ee8f63-a257-4210-8574-ed618b243d45",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.04,
        "w": 0.45,
        "x": 0.03,
        "y": 0.14
      },
      "kind": "title",
      "text": "Speech recognition finds its commercial feet",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "606ddfd3-848e-41d9-beca-e97747670132",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [
    {
      "name": "List presentation",
      "slug": "list-presentation",
      "agent": null,
      "layer": "slide",
      "matchId": "a82a4e9d-4229-4165-8ef4-f40e6c11ce35",
      "evidence": "The slide contains multiple list/bullet components with several items.",
      "confidence": 0.8
    }
  ],
  "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": 174,
      "from": 86,
      "beatId": "019dd95a-0682-776c-8e35-73133c63a096",
      "arcName": "Voyage and Return",
      "arcSlug": "voyage-return",
      "beatName": "Discoveries",
      "beatSlug": "voyage-return-discoveries",
      "evidence": "Industry + Politics findings on NVIDIA, regulation, geopolitics",
      "position": 3,
      "confidence": 55,
      "parentBeatName": "Development",
      "parentBeatSlug": "development"
    }
  ],
  "loops": [
    {
      "to": 136,
      "from": 131,
      "name": "Pattern Hunter",
      "slug": "02-pattern-hunter",
      "bestFor": "Time-pressed audiences, building consensus, when data is strong",
      "matchId": "019dd95a-07fe-70ce-8d3e-5814a3b51188",
      "evidence": "ElevenLabs $1.1B, GenAI app growth, speech-recognition, GPT-4o/Moshi voice",
      "position": 15,
      "objective": "Document the rise of speech and audio interfaces",
      "structure": "Evidence A -> Evidence B -> Evidence C -> Pattern/Conclusion",
      "confidence": 70,
      "description": "Group multiple pieces of evidence that together point to a pattern or conclusion"
    }
  ],
  "imagePathAlt": null,
  "thumbSrc": null,
  "thumbSrcAlt": null,
  "locked": true
}