🦅🐼 Sizing up US-China AI strategy: A Quick Q&A with … Kyle Chan

'With AI, I think there’s too much zero-sum framing.'

My fellow pro-growth/progress/abundance Up Wingers,

The race for AI dominance is a high-stakes feature of the US-China rivalry. While both countries pursue leadership in the field, their priorities may diverge on issues like superintelligence vs. practical application. Likewise, their different areas of strength lend themselves to different approaches. I asked Kyle Chan a few quick questions about how to track real progress, what strategies each side is pursuing, and the mistakes the US can’t afford to make.

Chan is a postdoctoral research associate and sociology lecturer at Princeton. His research centers on industrial policy, clean tech, and infrastructure, particularly in China and India. Chan has contributed to several major media publications, including the recent New York Times op-ed, “In the Future, China Will Be Dominant. The U.S. Will Be Irrelevant.” He is also the author of the High Capacity Substack.

---

US compute is an order or two of magnitude greater than in China, and I think that gap could very likely grow over time . . .

1/ Is it useful to frame US-China competition in AI as a race, which suggests a zero-sum outcome?

Yes and no. On the one hand, I do think it is important, in general, to understand the role of nation states involved in the US-China tech competition more broadly, and then that includes AI. So it would matter a lot. For example, if all the leading tech companies in the world — in biotech, in AI, in material sciences, in medical diagnostic devices — if those were all Chinese, it would change a lot of things geopolitically.

On the other hand, I think it can be overdone. In particular in AI, I think there’s too much of this sort of zero-sum framing. My approach to understanding this topic is more along the lines of the internet or computers and other similar general purpose technologies. Less of this sort of threshold or takeoff point and more about, more broadly: In a longer-term economic and technological competition, how can we push forward diffusion? How can we push forward model performance? How can we do all these things? It doesn’t always have to be that if we don’t get there and China gets there first, that’ll be the end of everything.