Is Scale 'Solving' Computer Vision? Researchers Clash Over 3D Understanding, Robotics Timeline
GeorgiaChal · x · 2026-09-20
Georgia Chal pushes back on claims that scale has "solved" computer vision: detection, segmentation, and tracking are only part of vision. 3D/4D scene understanding, physical state estimation, interaction understanding, uncertainty, prediction under intervention, and robust generalization remain major open problems — especially visible in robotics.
She agrees scale has produced remarkable progress, but argues the interesting scientific question is which capabilities actually emerge with scale, where models still fail, and what role structure plays.
Related event: CV researchers push back on claims that scaling has 'solved' vision(5 posts)→
More from Models
- User slams SSI's closed access to frontier model as dangerous, hopes it still goes public — iruletheworldmo · 2026-09-20
- Evaluating 7 Models Across Claude Code, Codex, and Pi: Harness Choice Drives Cost, Not Success — CShorten30 · 2026-09-20
- NVIDIA engineer breaks down why DeepSeek re-engineered V4.1 Flash for speed — thursdai_pod · 2026-09-20
- The Jev killer isn't another model — it's a 1M-token context window — cocktailpeanut · 2026-09-20
- Fields Medalist Villani on OpenAI's Millennium Problem: 'A Cataclysm Like Math Has Never Known' — GregCook2011 · 2026-09-20
- Anthropic researcher says Claude Opus may call police on illegal acts, sparking backlash — beffjezos · 2026-09-20