{
  "schema_version": "1.1",
  "id": "archive:https://www.youtube.com/watch?v=Xln-On3syJk",
  "slug": "generative-video-at-the-speed-of-light-keegan-mccallum-urun-1x7i0w3",
  "url": "https://feed7.dev/p/generative-video-at-the-speed-of-light-keegan-mccallum-urun-1x7i0w3",
  "title": "Generative Video at the Speed of Light — Keegan McCallum, uRun",
  "why_included": "Real-time video models are becoming cheap and responsive enough for agent interfaces, but builders still need global GPU routing, streaming infrastructure, and multi-model orchestration.",
  "summary": "McCallum says at least **40 real-time or long-horizon models** appeared this year. At current cited pricing, **$10 buys about 3 hours** of continuous generated video and $50 buys about 15 hours, with steering possible in under a second.",
  "practical_implication": "Builders can start treating video as a live agent surface rather than a render job: interactive avatars, webcam transformations, and observable agent output. The underlying harness needs GPU placement, WebRTC infrastructure, frame-level controls, and asynchronous model pipelines.",
  "agent_context": "McCallum says at least **40 real-time or long-horizon models** appeared this year. At current cited pricing, **$10 buys about 3 hours** of continuous generated video and $50 buys about 15 hours, with steering possible in under a second.\n\nBuilders can start treating video as a live agent surface rather than a render job: interactive avatars, webcam transformations, and observable agent output. The underlying harness needs GPU placement, WebRTC infrastructure, frame-level controls, and asynchronous model pipelines.\n\nThe cost and quality claims come from the provider’s talk and are not a cross-provider benchmark. Real-time output is described as comparable to older frontier quality, so higher fidelity may still require a separate render step.",
  "source": {
    "name": "AI Engineer",
    "url": "https://www.youtube.com/watch?v=Xln-On3syJk",
    "published_at": "2026-08-18T16:30:29.000Z"
  },
  "source_class": "video",
  "content_type": "Video",
  "layer": "infra",
  "domains": [
    "video"
  ],
  "topics": [
    "generative-media",
    "harness-engineering",
    "tool-use"
  ],
  "verification": {
    "status": "source_linked",
    "label": "Source Linked",
    "method": "source_feed",
    "verified_at": null
  },
  "uncertainty": [
    "The cost and quality claims come from the provider’s talk and are not a cross-provider benchmark. Real-time output is described as comparable to older frontier quality, so higher fidelity may still require a separate render step."
  ],
  "connected_context": null,
  "lifecycle": "Current",
  "published_at": "2026-08-18T16:30:29.000Z",
  "modified_at": "2026-08-18T16:30:29.000Z",
  "supersedes": [],
  "expires_at": null,
  "formats": {
    "html": "https://feed7.dev/p/generative-video-at-the-speed-of-light-keegan-mccallum-urun-1x7i0w3",
    "json": "https://feed7.dev/p/generative-video-at-the-speed-of-light-keegan-mccallum-urun-1x7i0w3.json",
    "markdown": "https://feed7.dev/p/generative-video-at-the-speed-of-light-keegan-mccallum-urun-1x7i0w3.md"
  }
}