I think it’s mostly right, maybe lacking perspective at some points, like the losses of Chinese companies are nothing compared to the losses of US companies.
And some funny bits
Even if the Chinese labs make little money themselves, they can force American competitors to cut prices and make it harder to recover the billions they spend training new models.
Even if the Chinese labs make little money themselves, they can force American competitors to cut prices and make it harder to recover the billions they spend training new models.
but business insider, that’s just good business
“do good business” is only rule #1 of capitalism tho. it’s superseded by rule #0, which is “be extremely racist”
did someone spend billions on vapourwareI think there are entirely different incentive models for Chinese companies, aren’t there? Like, aren’t they getting paid by the government to develop these models and that’s how they’re making their money?
I don’t know about that. From what I can tell a lot if funded by Alibaba and Tencent, some by banks/hedge funds. If you have other information please share.
in the sense it’s open weights, but not the training data? yeah, i guess?
you know who else does it though? us “open” models like zucc’s llama. or pretty much all of them.
it’s not open-source in the sense that they’re not gonna do this thing that’s completely infeasible to do and wouldn’t provide all that many benefits anyway, unless they got every ai company to train off the same data instead of DDoSing every server they can find
tbf not ddosing everyone would be nice. i’m sure we could cooperate into a big repository of knowledge with an easy way to download and use for various such purposes if our society was a little healthier. the internet had promised us so much.
Free marketers when the free market free markets
