What are we actually seeing with GPT-6 Astra? A skeptic's look at the robotics results
GeorgiaChal · x · 2026-09-19
After digging into GPT-6 Astra's emerging results in robotics, 3D, CAD and autonomous experimentation, Georgia Chal wrote a blog post. She started skeptical and still doesn't think these results mean scaling has "suddenly solved robotics" — but something interesting is happening.
Her core questions:
- What capability are we actually seeing?
- What role does structure play?
- What is learned by the model itself, and what becomes possible only because of surrounding tools, representations and algorithms?
- Most importantly: what experiments do we actually need to answer these questions?
It's a methodological critique focused on building the right measurement framework, not picking sides between "scaling wins" and "demo hype".
More from Embodied
- DISCO distilled tree search named CoRL 2026 best-paper finalist, top 6 of 687 — yining_hong · 2026-11-04
- Berkeley's Value Forcing: enforcing Bellman consistency at every noise level for accurate RL value functions — berkeley_ai · 2026-11-04
- 22 Creative SAM 3.1 Use Cases From Meta's AI Developer Hackathon, From Wildfire Detection to Document Reassembly — Vjeux · 2026-11-04
- Ex-OpenAI robotics intern breaks down viral Astra robot-painting demo on Cerebras podcast — AccBalanced · 2026-11-04
- Embodied Agent Arena: Closed-Weight VLMs Lead All Five Robot Task Domains, Open-Weight Models Lag — Haojian Huang · 2026-11-04
- PrimeBot's PrimeSim: a generalist world model simulator trained on cross-embodiment demos — Travis Davies · 2026-11-04