Interesting take-away from the article for me is that there is nothing structural or cultural that would dictate Chinese labs will always be fast followers.

- a broader collaborative AI ecosystem in China vs warring labs in the US

- less “superstar” self promotion vs US

If there was unconstrained availability of Nvidia GPUs I could easily see the situation where the Chinese labs are a few months ahead of the US.

[dead]