2 ms·
I don't think anyone who's paid attention to the LLM scene will give them any “credit for catching up fast” as it has been pretty obvious for the past two years
by littlestymaar 2y ago
I don't think anyone who's paid attention to the LLM scene will give them any “credit for catching up fast” as it has been pretty obvious for the past two years that all it takes to reach the state of the art is a big enough GPU cluster.
Deepseek made the news because how they were able to do it with significantly less hardware than their American counterparts, but given that Musk has spent the last two years telling everyone how he was building the biggest AI cluster ever, it's no surprise that they manage to reproduce the kind of performances other players are showing.
- dmix 2y agoThis severely underestimates the talent still required. Deepseek didn't come out just because it's cheaper, it came out because a very talented team figured out how to make it cheaper.
- littlestymaar 2y agoFor Deepseek, I'm not saying otherwise, quite the opposite. But Grok hasn't shown anything that suppose the level of talent that Deepseek exhibited.
- jp42 2y agoeven if we assume you are comment is correct. Lets extrapolate what happens next. talented team, biggest compute among all competitors and CEO who is hell bent on winning the race. imo that is the reason it is big deal.
- littlestymaar 2y agoGrok (unlike Deepseek) has yet to show any ability to make conceptual breakthrough. I don't like OpenAI at all but one must admit that they are at least showing that they can move the field forward.