Tit for tat. US administration thought they would contain China. Now Deepseek v4 is running on their home grown chips. US companies lost that hardware revenue and dependency instead.

▲

Levitz 3 hours ago | parent | next [-]

Weird, why does China want GPUs then?

Also, regarding that "tit for tat", how large of a "tat" is due, regarding decades of corporate espionage?

▲

dyauspitr 4 hours ago | parent | prev [-]

Deepseek v4 is still pretty far behind the frontier models though.

	▲	BobbyJo 3 hours ago \| parent [-]
		It's really hard to tell. Almost all the models have the benchmarks in their training data, which pushes us into the realm of basing model capability rankings on vibes. I think the OSS models tend to do worse on things outside their corpus, but Deepseek specifically has done insanely good work on efficiency and scaling, which is verifiable in a way capabilities are not.