Why this book
Kai-Fu Lee spent his career on both sides of the Pacific — training in AI at Carnegie Mellon, running research at Apple and Microsoft, then leading Google's China operation before founding the venture firm Sinovation Ventures — and AI Superpowers is his argument that this vantage point reveals something Silicon Valley-centric coverage of AI misses. Once deep learning matured around the mid-2010s, Lee contends, the AI race stopped being primarily about who has the cleverest researchers and started being about who has the most data and the most aggressive entrepreneurs willing to grind out real-world deployment. On that second criterion, he argues, China's market scale, lighter data-privacy constraints, and famously combative startup culture give it a genuine shot at parity with or advantage over the United States — not because Chinese AI is more inventive, but because implementation, not invention, is where the coming decade's value gets captured.
The book's second half turns from geopolitics to consequence: Lee predicts that AI-driven automation will displace a much larger swath of white-collar and blue-collar work than most experts of the period expected, and he's skeptical that universal basic income alone addresses the resulting need for meaning, not just money. He closes with a personal turn — a cancer diagnosis that reshaped his sense of what human work should be for — proposing that societies redirect AI-generated wealth toward care work and human connection that machines can't replicate. Written in 2018, the book's specific predictions about the pace and balance of the US-China AI race have since been tested by events Lee couldn't foresee — export controls, shifts in open-source models, and swings in each country's execution — so its forecasts are best read as an informed argument of their moment rather than settled fact.
Who should read it
Readers wanting a China-side view of the AI competition, unfiltered by Silicon Valley's usual framing, will find real value here, as will anyone trying to understand why deployment speed can matter as much as research breakthroughs. It will frustrate readers seeking rigorous economic modeling of job displacement, or anyone uncomfortable with a framework that treats the US-China dynamic in fairly zero-sum, national-competition terms — a framing some critics argue overstates rivalry at the expense of more nuanced analysis.
About the author
Kai-Fu Lee is a Taiwan-born, U.S.-educated computer scientist who earned his PhD in AI from Carnegie Mellon University and held executive roles at Apple, SGI, Microsoft, and Google, where he served as founding president of Google China. He is the chairman and CEO of Sinovation Ventures, a venture capital firm investing in Chinese technology startups.