SoftBank, Sony, Honda, and six other Japanese companies have formed a new AI company to develop a massive foundation model for physical AI by 2030. The 1 trillion-parameter model represents Japan's bid to compete in advanced AI development.
Nine major Japanese corporations including NEC, Fujitsu, and Sumitomo Chemical have joined forces to establish the venture. The collaboration targets creating a 1 trillion-parameter foundation model—roughly 10 times larger than current leading models—specifically designed for physical AI applications.
Physical AI refers to artificial intelligence systems capable of interacting with and manipulating the physical world through robotics and autonomous systems. The focus reflects growing demand for AI that extends beyond language and text processing into real-world industrial and manufacturing applications.
SoftBank leads the initiative, positioning Japan to develop domestic AI capabilities rather than relying solely on foreign technology. The 2030 deadline provides a decade-long development window for the consortium to achieve the ambitious technical and computational goals.
The partnership pools resources and expertise across multiple sectors—financial services, consumer electronics, automotive, and industrial manufacturing. This cross-industry approach aims to address diverse use cases from robotics to industrial automation.
The move signals Japan's strategic response to AI leadership by the United States and China. By combining institutional resources, the consortium attempts to reduce individual company costs while accelerating development timelines. The venture also represents an effort to establish Japan as a significant player in foundational AI model development, a field currently dominated by American companies.
Details regarding funding, organizational structure, and specific technical roadmaps remain unclear. The announcement comes as global competition intensifies around frontier AI model development, with major tech companies racing to scale model parameters and capabilities.
Japan's approach differs from the typical U.S. model of individual companies competing separately. The consortium strategy emphasizes collective national competitiveness in AI infrastructure and capability development.
Apollo chief economist Torsten Slok warns that AI-driven profit gains in non-tech sectors could arrive significantly later than Wall Street expects, potentially triggering major stock repricing.
A British space startup has deployed a longevity research laboratory into orbit. The facility will transmit data to train AI models capable of predicting protein behavior linked to age-related diseases.
UN Secretary-General António Guterres called for lethal autonomous weapons systems to be banned under international law, labeling them "morally repugnant." The push highlights a growing policy divide between the US Department of Defense and AI company Anthropic.
An OpenAI artificial intelligence model has successfully disproven a longstanding conjecture in discrete geometry, marking a significant breakthrough in using machine learning to solve fundamental mathematical problems.