investors don’t care about the legality of something as minor as copyright law when those same investors work in the government themselves and have deep state ties, corporate America is especially corrupt in such cases.
So hypothetically if you wrote a book and it sold relatively well and then Open AI used it to train it’s model using your book with any compensation or permission, you wouldn’t care about it? And what about when I asked it to quote directly from the book and it the results were reproduced verbatim? It’s not even about it being reproduce verbatim or them providing a summary or whatever, it is about the fact that they did so without explicitly stating that what all data they had collected and trained upon and used it to profit themselves. It’s not always about the gray legal areas but it is also about the transparency.
No actually it is far worse since in text they can still cite the original source but when using their Sora model there is no clear indication as to what the produced result was trained on and that is where it gets dangerously close to blatant copyright infringement.
And why should Deepseek? Again what they did nothing different than what Open AI did, they just stole from a singular multi billion dollar company and Open AI stole from millions of smaller entities.
I don’t care about Open AI using the works of others to train AI much myself but when they are put in a special box and treated as if they are somewhat special and different than other, well that’s where I have a problem.
Not just AI but particularly Open AI and what is going on behind the scenes.
And when saying that “Open AI did their own collection and them stealing is different than Deepseek doing it” you also have to consider that both of them started from a separate baseline, had the roles been reversed Open AI would have 1000% did the same.