{
  "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": 27,
  "pageCount": 136,
  "prevPage": 26,
  "nextPage": 28,
  "slideType": "industry_trends",
  "function": "present_solution",
  "density": "dense",
  "nDataPoints": 0,
  "notes": "The diagram shows three layers of ML systems (Generic, Automotive, Network) and their data flow from physical domain to digital representation to model application.",
  "elementsJson": [
    "headline_text",
    "bullet_list",
    "process_diagram"
  ],
  "metadataConfidence": 0.95,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5c-f6fdd8f57895/27",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5c-f6fdd8f57895",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5c-f6fdd8f57895.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5c-f6fdd8f57895#slide-27",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": "The attack surface of ML systems is large: Adversaries can manipulate data collection, corrupt the model or tamper with its outputs.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-41ab-72d9-bbc3-553714ff5f11",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.45,
        "w": 0.58,
        "x": 0.4,
        "y": 0.35
      },
      "kind": "diagram",
      "text": "Diagram showing data flow through ML systems across three domains: Generic, Automotive, and Network.",
      "attrs": null,
      "subkind": "process",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "f1228f13-0be9-4d19-aaa3-82408d9baaf7",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.25,
        "w": 0.42,
        "x": 0.03,
        "y": 0.32
      },
      "kind": "list",
      "text": "TensorFlow Privacy from Google allows for training ML models on users' data while giving strong mathematical guarantees that they do not learn or remember details about any specific user. The library is also designed to work with training in a federated context.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "a5ff6fc8-2016-4dce-9029-98f7a096e159",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.15,
        "w": 0.42,
        "x": 0.03,
        "y": 0.71
      },
      "kind": "list",
      "text": "TF Encrypted from Dropout Labs is a library built on top of TensorFlow to integrate privacy-preserving technology into pre-existing ML processes.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "f9e284a4-96af-4647-a3db-00f83bb4c4a1",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.08,
        "w": 0.42,
        "x": 0.03,
        "y": 0.22
      },
      "kind": "paragraph",
      "text": "The attack surface of ML systems is large: Adversaries can manipulate data collection, corrupt the model or tamper with its outputs.",
      "attrs": null,
      "subkind": "paragraph",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "b8d1215b-03e2-467e-8124-729e809257a3",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.05,
        "w": 0.94,
        "x": 0.03,
        "y": 0.13
      },
      "kind": "title",
      "text": "Increasing emphasis on data privacy and protecting deployed ML systems from attacks",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "ef8a15f7-df28-496d-9d82-9eeaf1eb6506",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [
    {
      "name": "Action Titles",
      "slug": "action-titles",
      "agent": "Architect",
      "layer": "slide",
      "matchId": "019dd95a-0fd5-7148-8ecc-df9a13de19ad",
      "evidence": "Callout names ML attack surface.",
      "confidence": 70
    }
  ],
  "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": [],
  "imagePathAlt": null,
  "thumbSrc": null,
  "thumbSrcAlt": null,
  "locked": true
}