On October 27, Zhiyuan officially launched Genie Master, an embodied agent framework. The Spirit G2 robot equipped with the framework has been working at a real convenience store without manual demonstrations or debugging, achieving continuous self-evolution with no human in the loop, and marking the first validation of Physical RSI (embodied recursive self-evolution) in a real business scenario.
At Unitree's 10th anniversary ecosystem conference, founder Wang Xingxing said the world still underestimates AI, identified data, computing power, chips and energy as the most critical future resources, and predicted physical AI could see key progress in the coming years. The company plans to expand robot production and increase research into AI models and computing.
Professor Li Yiming's team released two Physical AI research results: the world action model PrismWAM, which ranked first in the RoboTwin 2.0 evaluation, and the agentic framework RoboDoctor, which scored 57.1 across all 54 tasks in RoboDojo, leading the current comparison.
Formic announced it raised $50 million in growth capital, including a Series B led by Mercato Partners and debt financing from S2G Investments, to accelerate its expansion into general-purpose robotics and new applications such as metal fabrication.
At GTC Berlin, Jensen Huang reiterated that "the next wave is physical AI" and said AI factory deployment can be shortened to 14 days at the fastest.
Unitree held its 2026 Physical AI Conference and Partner Ecosystem Conference, where Wang Xingxing delivered a keynote speech and showcased new products including the G1+ and H2i humanoid robots, Dex5-S dexterous hand, and R1-7a robotic arm, with over 40 exhibition areas.
Techable Ventures founding partner Amulya Vishwanath said in an interview that physical AI is in a "LEGO era," with robotics' foundational software and building blocks being rethought, and that talent competition with large companies like Anthropic and Nvidia is key; her firm helps founders connect with top talent.
Mistral AI announced a new collaboration with Nvidia to bring AI to industrial engineering, including faster simulation, open physics AI benchmarks with supporting datasets, and evaluations for agentic AI in CAE and EDA.
NVIDIA announced SimReady Central Library, a free collection of over 2,000 simulation-ready OpenUSD assets on Hugging Face for physical AI training and AI factory digital twins, covering robot arms, grippers, warehouse equipment, racks, and cooling infrastructure, with OEMs able to validate and distribute assets to developers.
Inverse Matrix, together with Beijing Academy of Artificial Intelligence, vLLM, and Peking University, released PhysisServe, the first physical AI inference system that balances intelligence and high-throughput output. It is the first to enable a 10-billion-parameter world model to achieve real-time kilohertz output frequency, one to two orders of magnitude faster than existing frontier solutions.