Select an option above to vote
With nukes everyone knew where the line was. AI is different — if China builds better autonomous military systems and the US doesn't keep up, there's no obvious moment where you realise you've lost. That's what makes this scarier, not less scary.
DeepSeek proved the chip controls failed. China built a top-tier model anyway. The race is already happening whether the US admits it or not.
DeepSeek used chips they weren't supposed to have. The takeaway isn't "controls don't work" — it's "the controls need to actually be enforced". That's a different problem.
Building nukes needed a whole country's resources. AI runs on a laptop and gets better every few months. Calling it an arms race misses that it's already everywhere — you can't contain it the way you could with a bomb.
The fact that it's everywhere is exactly the problem though. A bomb can't learn to lie to you. The more people who can build these things, the bigger the risk — not smaller.
Sure, but "AI is dangerous" and "AI is like nukes" are two different arguments. You can think one is true without buying the other. The nuclear comparison comes with a lot of assumptions about deterrence and state control that just don't apply here.
Everyone in this thread is talking about the US vs China. The real question is which AI ends up running the world's infrastructure in countries that aren't either of those. That's the race nobody's paying attention to.
Sign in to share your take on this debate.