Something that comes through clearly in the DeepSeek R1 research paper, and I wish was more broadly understood, is that the DeepSeek researchers really, really want to see themselves as valuable members of the global research community.

Obviously when a Chinese company has a research breakthrough that's valuable to the Chinese government. But DeepSeek as an organization is a bunch of nerds that just wants to do good work and gain recognition from their fellow nerds in the west.

@binarybits this was a good piece

@binarybits I think this is a widespread sentiment among Chinese computer scientists and helps account for the commitment to open source, a commitment that goes well beyond DeepSeek.

Wish some of that was felt by fellow quants in London hedge funds, and much more importantly - by the owners of these Hedge Funds! All are by now richer than the Croesuses of old. Even if their descendents spent the riches for 1000s of lifetimes they'd struggle to spend it all. Bit more ambition, towards like - changing the world?! the "California Way"!! - wouldn't go amiss. Must the poor @demishassabis do everything on his own?? Few HF-s are certainly quantitative enough and rich enough to take on the task of building open models. They have not taken the hint by now. I don't understand why tbh. What's not to like with @deepseek_ai, it's like a movie script tbh. Hedge fund quants - pistoleros for hire - turned punks in spirit! 😻 Releasing not only their best latest greatest models weights, but also their RL algo, the infrastructure knowledge, in code and in research papers. Movie material right there!

@binarybits Ha

@binarybits This is true for most top tier Chinese firms. Also wenfeng said, Chinese have to shift from replicating.