DeepMind's Jakob Engel on World Models: Egocentric Multimodal Data Is the Future
Google DeepMind researcher Jakob Engel argued that future foundation world models will be built on and oriented toward egocentric, multi-view, multimodal data. The discussion was sparked by @duchaaiki's observation that humans and VLMs produce strikingly similar outputs when reconstructing images from memory or text descriptions.
2026-09-09 ~ 2026-09-09 · 4 related posts
- DeepMind's Jakob Engel shares his take on world models — ducha_aiki · 2026-09-09
- Humans and VLMs produce surprisingly similar image reconstructions from memory — ducha_aiki · 2026-09-09
- Follow-up link on humans and VLMs producing similar memory-based reconstructions — ducha_aiki · 2026-09-09
- Future foundation models will be built on egocentric, calibrated multiview data — ducha_aiki · 2026-09-09