GPT-6 Astra hints robots may inherit the scaling curve of general intelligence
liuziwei7 · x · 2026-09-10
GPT-6 Astra didn't solve embodied tasks, but Kuvvius argues it points to something bigger: robotics may inherit the scaling curve of general intelligence as a downstream capability.
- This month's signals: Google DeepMind's Gemini Robotics 2 placed an embodied reasoning model above a VLA for whole-body control; Anthropic tested general-purpose reasoning models at several levels of control
- The bad news: general reasoning models still cannot control a humanoid on their own — a pretrained policy is needed underneath
- The bet: the winning model may not start as a specialized robot model, but a general-purpose one that understands the physical world and closes the loop into body control — a robot body as just another interface for a general agent, like a computer but much harder
HumanCLAW is designed to measure exactly this trajectory.
Related event: GPT-6 Astra Dominates Embodied Benchmark, Nearly Tripling Success Rate(2 posts)→
More from AGI Musings
- Aeon's former EVP: AI agents are the most impactful innovation of my lifetime — akumar · 2026-10-27
- CJR: AI can't really 'go rogue' — why journalists should stop anthropomorphizing agents — ShakeelHashim · 2026-10-27
- Richard Ngo: the real question is whether we keep Christianity's metaphysical baggage — RichardMCNgo · 2026-10-27
- Physicists Let AI Pick Its Own Problem—It Wrote a Publishable Paper in 2 Hours — burny_tech · 2026-10-27
- Deep learning veteran warns: don't let entrenched disbelief derail prep for the next big tech — jachiam0 · 2026-10-27
- Should platforms serve third-party AI agents? The antitrust question — sjgadler · 2026-10-27