World models can 'see into' any drawing, photo or render: a computer graphics renaissance
XRarchitect · x · 2026-11-06
Quoting @neilsonks: we are in a computer graphics renaissance — drawings, photos, and renders are all things world models can now "see into", producing what he calls "world vertigo". The post links to a demo video.
The observation highlights the leap in world models' ability to turn any static visual artifact into an explorable world.
More from AGI Musings
- Teknium pushes back on METR's 'no bottlenecks' claim, citing funding and compute cost inflation — Teknium · 2026-11-06
- If You're Compute-Constrained, Don't Burn Agent-Compute on Social Agents — nptacek · 2026-11-06
- Does Anthropic's staff really believe in superintelligence? Insiders doubt culture scales — austinc3301 · 2026-11-06
- Haider: the 'one superpower wins the AI race' narrative is getting absurd — haider1 · 2026-11-06
- AI Engineer SH talk: models see the world through their action space, not ours — vincent_koc · 2026-11-06
- ThursdAI pod: midterms, 40 datacenter ballot measures, $3T wipeout — is AI over? — altryne · 2026-11-06