I actually agree with Jensen that export controls have largely backfired — but his talking points feel a bit disingenuous. (Of course, he’s lobbying for outcomes that protect a $5T company; exaggeration is part of the playbook.) The numbers Jensen cites don’t add up unless he’s defining “AI” very narrowly for Silicon Valley and very broadly for China. Chinese AI companies today are leaner because they actually .... also use AI. (... That said, there may indeed be a 50x gap in robotics talent if you count the entire ecosystem. I haven't done a real count but feels directionally correct. I think this doesn't get nearly enough attention.) Also, it wasn’t export controls that “mobilized” Chinese AI entrepreneurs. It was AI itself (and specifically ChatGPT). The lightbulb moment was: “AI is real, it’s massive, and we’d be crazy not to build it.” There’s some “China can do it too” energy, sure, but it’s mostly independent conviction. (That's not to say the sanctions did not speed up semiconductor industry buildout -- it absolutely did, but that's ... not the same industry and not the numbers being cited here)