{
  "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": 53,
  "pageCount": 163,
  "prevPage": 52,
  "nextPage": 54,
  "slideType": "industry_trends",
  "function": "summarize",
  "density": "dense",
  "nDataPoints": 0,
  "notes": "The slide highlights specific research papers and models (DreamFusion, RealFusion, SKED, Instruct-NeRF2NeRF) in the context of 3D generative AI.",
  "elementsJson": [
    "headline_text",
    "paragraph",
    "bullet_list",
    "photo"
  ],
  "metadataConfidence": 0.95,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5d-06b04d219fea/53",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5d-06b04d219fea",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5d-06b04d219fea.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5d-06b04d219fea#slide-53",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": "NeRF-based generative models are a promising direction for large scale creation of 3D assets.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-47c4-7719-a227-540d4afd4cce",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.25,
        "w": 0.96,
        "x": 0.02,
        "y": 0.7
      },
      "kind": "image",
      "text": "Visual examples of 3D generation and editing results.",
      "attrs": null,
      "subkind": "illustration",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "e5a5b317-e15b-4b76-ae0e-da303a65641c",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.35,
        "w": 0.96,
        "x": 0.02,
        "y": 0.33
      },
      "kind": "list",
      "text": "DreamFusion and Score Jacobian Chaining were the first methods to use a pretrained 2D text-to-image diffusion model to perform text-to-3D synthesis. Early attempts showed cartoonish-looking 3D models of single objects.\nRealFusion finetunes the diffusion prior on a specific image to increase that image’s likelihood.\nSKED only alters a selected region of a NeRF provided through a few guiding sketches. They preserve the quality of the base NeRF and ensure that the edited region respects the semantics of a text prompt.\nInstruct-Nerf2Nerf edits an entire NeRF scene rather than a region or generating from scratch. They apply a latent diffusion model on each input image and iteratively update the NeRF scene ensuring it stays consistent.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "02e7dad5-76a6-4954-a6eb-89c2c6af64f7",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.1,
        "w": 0.96,
        "x": 0.02,
        "y": 0.21
      },
      "kind": "paragraph",
      "text": "NeRF-based generative models are a promising direction for large scale creation of 3D assets. NeRFs not only have improved in speed and quality (see HyperDiffusion, MobileNeRF, Neurolangelo and DynIBAR) but also enabled GenAI to model 3D geometry.",
      "attrs": null,
      "subkind": "paragraph",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "3ceddcfc-a10e-4920-ba4e-ae67c3a2fdc7",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.04,
        "w": 0.2,
        "x": 0.02,
        "y": 0.14
      },
      "kind": "title",
      "text": "NeRFs meet GenAI",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "865291a6-39bd-497d-b7ae-3a95a94979fc",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [
    {
      "name": "Storytelling: audience calibration",
      "slug": "audience",
      "agent": "storyteller",
      "layer": "slide",
      "matchId": "64713e39-a3d2-48ad-a1f4-5362f31fbe1c",
      "evidence": "NeRF-based generative models are a promising direction for large scale creation of 3D assets.",
      "confidence": 0.7
    }
  ],
  "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": 58,
      "from": 50,
      "name": "Pattern Hunter",
      "slug": "02-pattern-hunter",
      "bestFor": "Time-pressed audiences, building consensus, when data is strong",
      "matchId": "019dd95a-07fe-70ce-8d3d-f70f6435b722",
      "evidence": "Text-to-video, instruction editing, Gaussian splatting, NeRFs, depth, SAM, DINOv2, weather, music.",
      "position": 6,
      "objective": "Generative media and visual foundation models proliferate",
      "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
}