China spends 42% more on robotics as AI race turns physical
The Multimodal Pivot
Focusing exclusively on generative AI and Large Language Models (LLMs) is a short-sighted trap. While the West remains fixated on digital chatbots, the global artificial intelligence race is rapidly becoming multimodal. This shift signifies a move from software that simply predicts text to hardware that navigates reality. The divergence in investment strategies between the two superpowers reveals a fundamental disagreement on where the ultimate economic value of AI resides.
China Outpaces U.S. in Physical Embodiment
Data from the private sector underscores a startling strategic divide. The United States currently outspends China by a factor of 12 in raw compute power. However, China has seized the lead in the robotics sector, spending 42% more than its American counterparts. This gap is not a temporary fluctuation; it is expected to widen as Beijing prioritizes the integration of intelligence into physical forms. China's advantage extends beyond mere assembly, encompassing sophisticated hardware and the specialized software required for mobility.
Defining the Physical AI Era
We are entering the era of physical AI. This is a categorical leap from traditional industrial automation. Standard factory machines are programmed for repetition within static environments. In contrast, physical AI involves machines that perceive, understand, and interact with an unpredictable physical world. These systems utilize multimodal AI to process visual, tactile, and spatial data simultaneously, allowing them to perform complex tasks that were previously the sole domain of human labor.
Implications for Global Manufacturing
China's dominance in physical AI threatens to rewire global supply chains. By mastering the intersection of robotics and AI, they are positioning themselves to automate sectors that the West has long considered too complex for machines. This is a play for the future of manufacturing and logistics. If the U.S. continues to focus its capital on compute for LLMs while neglecting the physical manifestation of that intelligence, it risks losing the foundational hardware race of the 21st century.
- Alice Han
- 13%· people
- artificial intelligence
- 13%· technologies
- China
- 13%· places
- James Kynge
- 13%· people
- Large Language Models
- 13%· technologies
- Other topics
- 38%
The AI race is about WAY MORE than LLMs, and now it's PHYSICAL
WatchThe Prof G Pod – Scott Galloway // 1:18
NYU Professor, best-selling author, business leader and serial entrepreneur Scott Galloway cuts through the biggest stories in tech, business, and investing with unfiltered insights, bold predictions and thoughtful advice. Podcasts include Prof G Markets with co-host Ed Elson, Prof G Conversations and Office Hours with Prof G.