[dead]
OpenAI says it has evidence DeepSeek used its model to train competitor
51–60 of 1001 posts
Re: OpenAI says it has evidence DeepSeek used its model to train competitor
#52As the article points out, they are arguing in court against the new york times that publicly available data is fair game. The questions I am keenly waiting to observe the answer to (because surely Sam's words are lies): how hard is OpenAI willing to double down on their contradictory positions? What mental gymnastics will they use? What power will back them up, how, and how far will that go?
Their way of squaring this circle has always been to whine about "AI safety". (the cultish doomsday shit, not actual harms from AI)
Sam Altman will proclaim that he alone is qualified to build AI and that everyone else should be tied down by regulation.
And it should always be said that this is, of course, utterly ridiculous. Sam Altman literally got fired over this, has an extensive reputation as a shitweasel, and OpenAI's constant flouting and breaking of rules and social norms indicates they CANNOT be trusted.
Re: OpenAI says it has evidence DeepSeek used its model to train competitor
#53It's reasonably likely that a lot of people linked to the federal government want to ban DeepSeek. You can tell it's being presented away from "they gave us a free set of weights" and towards "they destroyed $1T of shareholder value." (By revealing that Microsoft et al. paid way too much to OpenAI et al. for technology that was actually easy to reinvent.)
Theoretically this should be good for OpenAI - in that they can reduce their costs by ~27x and pass that along to end users to get more adoption and more profit.
Re: OpenAI says it has evidence DeepSeek used its model to train competitor
#54OpenAI is also possibly in violation of many IP laws by scraping the entirety of the internet and using to train their models, so there’s that.
Re: OpenAI says it has evidence DeepSeek used its model to train competitor
#55"You can't take data without asking" seems like a court precedent OpenAI really, really, really wants to avoid. And yet...
Why? When did large companies care about laws? See e.g. Uber, AirBnb. The only thing government cares about at this point is if information is shared with China.
They care when they get big enough to attract attention from people like state AGs who can actually put the hurt on a bit. Uber and AirBnB both hit this point years ago; OpenAI's starting to hit it.
Re: OpenAI says it has evidence DeepSeek used its model to train competitor
#56It's reasonably likely that a lot of people linked to the federal government want to ban DeepSeek. You can tell it's being presented away from "they gave us a free set of weights" and towards "they destroyed $1T of shareholder value." (By revealing that Microsoft et al. paid way too much to OpenAI et al. for technology that was actually easy to reinvent.)
Would it even matter? Isn't the cat out of the bag and everything they did repeatable by an American research team?
American researchers had already made enough progress to prove that LLMs were not an incomprehensible trade secret based on years of secret knowledge - investors and tech executives were simply lead to believe otherwise. Well-connected people are probably very mad about this and they may try to lash out like the emotional human beings they are.
Re: OpenAI says it has evidence DeepSeek used its model to train competitor
#57It's reasonably likely that a lot of people linked to the federal government want to ban DeepSeek. You can tell it's being presented away from "they gave us a free set of weights" and towards "they destroyed $1T of shareholder value." (By revealing that Microsoft et al. paid way too much to OpenAI et al. for technology that was actually easy to reinvent.)
Theoretically this should be good for OpenAI - in that they can reduce their costs by ~27x and pass that along to end users to get more adoption and more profit.
No; those costs were their moat.
Re: OpenAI says it has evidence DeepSeek used its model to train competitor
#58It's like that Dr Phil episode where he meets the guy who created Bum Fights!
Re: OpenAI says it has evidence DeepSeek used its model to train competitor
#59While I'm as amused as everyone else - I think it's technically accurate to point out that the "we trained it for $6 mio" narrative is contingent on the done investment by others.
OpenAI's models were also trained on billions of dollars of "free" labor that produced the content that it was trained on.
Re: OpenAI says it has evidence DeepSeek used its model to train competitor
#60There is an Egyptian say that would translate to something like
"We didn’t see them when they were stealing, we saw them when they were fighting over what was stolen"
That describes this situation. Although to be honest all this aggressive scraping is noticeable but for people who understand that which is not majority of people. but now everyone knows.