Comment by lorecore
1 day ago
It’s certainly too early to call (if you must view this as some sort of adversarial competition). The US is behind on local models, the future for anyone who cares about privacy. There may be step change innovation yet to come that completely shifts the landscape. There’s basically no switching costs to users to change models. They have no lock-in.
It is very much adversarial, and to view it anything but adversarial is to not see the geopolitical reality and the potential national security implications of AI for what they actually are. Moreover, to claim that China is in the lead with local models presupposes that openai and anthropic could not release local models that are better, which is a big assumption. They do not release such models because they have frontier-grade propietary models that have high value.
As someone who happens to have been born in the US and currently lives here, I welcome China winning. I trust them infinitely more than I trust my own government and industry.
OpenAI and Anthropic are beholden to the capitalist system they exist under and hence cannot compete on local models. Like you say, they must try to maximize shareholder value. China is unencumbered by that constraint.
What is it about China that makes you trust them more?
8 replies →
You are free to hate capitalism (even if you benefit from it enormously). You are free to say that you hate capitalism and the U.S. as openly and as often as you like, without facing imprisonment or worse.
But if you were in China, could you say you hate the Chinese Communist Party and China openly and as often as you like without imprisonment or worse?
We know the answer to that. So go ahead and trust China more than the U.S., but I think that is pure foolishness.
7 replies →