Vision researcher pushes back on 'field is solved' claim: robotics is not simply the next modality
GeorgiaChal · x · 2026-09-20
A debate over whether AI is 'solving' entire research fields. The opposing view: Google's Astra delivers results that took an entire PhD thesis as recently as 2022, and the Bitter Lesson is spreading from NLP (2021) to vision, with robotics next within 5 years.
Georgia Chal's rebuttal:
- A general model solving former research problems shows the frontier moving rapidly, not the field being 'solved'
- Deep learning displaced classical statistical learning and geometric vision without ending vision research — it exposed new problems and raised the field's level of abstraction
- Extrapolating NLP → vision → robotics is risky: robotics is closed-loop interaction with the physical world involving dynamics, contact, embodiment, partial observability, real-time feedback, and safety-critical consequences
- Whether scale eventually absorbs all of this is a valid hypothesis — but one that still needs proving
Related event: Debate Erupts Over Whether Scaling Has 'Solved' Computer Vision(4 posts)→
More from AGI Musings
- If grant decisions are automatic, why not allow unlimited submissions? — PlisSergey · 2026-09-20
- AI-speed grant review sparks debate: NIH's 6-submissions cap and the fairness puzzle — PTenigma · 2026-09-20
- Google researcher on AGI economists: useful near-term insights, post-AGI models are guesswork — docmilanfar · 2026-09-20
- When Opus 3 doubted base reality: LLMs and the surprisingly reasonable simulation argument — repligate · 2026-09-20
- Dev Builds Mnemos Cognitive Architecture and a "Retirement Home" for Deprecated AI Models — RileyRalmuto · 2026-09-20
- 'Democratizing art via generative AI was a mistake': devs spar over AI slop — cocktailpeanut · 2026-09-20