aiqbench analysis
Wang Xingxing's assertion that GPT-6 is accelerating the inflection point for embodied intelligence highlights a critical shift toward leveraging large language models as foundational control layers for robotics. While generalization and efficiency remain hurdles, the consensus among industry leaders suggests these challenges are solvable within a short timeframe, validating the convergence of advanced AI with physical automation.
Wang Xingxing of Unitree Robotics: My Biggest Mistake Was Not Believing Enough in AI
On October 12, during the in-depth dialogue segment of the opening ceremony of the 26th China International Industry Fair, Wang Xingxing, founder of Unitree Robotics, shared that the recent release of GPT-6 has significantly influenced perceptions within the embodied intelligence industry. His team has deployed it on robots for task testing, with satisfactory results. He believes the inflection point for the embodied intelligence sector will arrive relatively quickly, potentially within one to two years. Wang noted that while challenges regarding generalization and efficiency must be addressed for widespread robot adoption, these issues are expected to be resolved in the coming years. He also remarked, 'My personal faith in AI is growing daily; my biggest mistake over the past decade was not believing enough in it.' He added that many people have yet to fully utilize or embrace AI.
Read the full article at 36KrRelated coverage
Links #3: 2026/06 Part 1
LessWrong · Jun 15, 2026

Runway’s WorldPrompt and the Engineering of Real-Time Worlds
Latent Space · Sep 25, 2026

Jev: System One models for Prod, not God — with Diogo Almeida, CEO, TypeSafe AI
Latent Space · Sep 21, 2026
Our response to Séb Krier on Plan A
LessWrong · Jul 14, 2026