Generative AI + Spatial Computing = ∞
summary
Claims that generative AI makes mass 3D UGC practical, demonstrated by a working prototype game called Triple Go where GPT + Tripo generate and fuse assets into playable creatures; pipeline auto-refines, re-topologizes and exports to USDZ in seconds costing roughly $1 per USDZ export, producing 10,000 models/day in production. Limitations include rough textures, high polygon counts and tricky rigging so not AAA-ready, but the demo on Vision Pro proves the idea: rapid asset creation drives new, curiosity-driven gameplay and lowers content costs.
AI-written summary · about
chapters (10)
- 0:00Introduction VR journey
- 3:53VR content challenges
- 7:47Why VR content failed
- 10:39AI revolution arrives
- 15:51AI 3D content potential
- 18:17Triple Go gameplay
- 21:26AI game development future
- 24:383D modeling cost drop
- 27:46Vision Pro UGC opportunity
- 32:42Call to creators
related talks
30mGenerative AI will enable crowdsourced interactivity
27mAI Pipeline for 3D Content: From Idea to 3D Generation
27mWorking Smart - Machine Learning in Art Asset Creation for the Games Industry
33mThe Dawn Of Universal Creativity
18mThink You Can Build a Game with AI? Think Again!
20mGenerative Design
39mGenerative Machines
28mIt's Natural to Be a Machine