A week ago I thought about how the old Copyleft of the left had become the new request of the Big Tech of the AI (Here the article).
Now, following a cause in the United States intent on Halfit comes out that to train its artificial intelligence, Llama 3, an archive of one million books under copyright has been used (the material was taken from Book3). Meta appeals to the “Fair Use”, already used by artists to produce works from works protected by copyright, but it is a bit of a climbing on the mirrors of digital competition (other companies, not to mention China).
I am reminded of a discussion that I made the other night with my friends Gigi Ballarani and Klara Murnau, especially with the second, convinced that traveling means knowing, so even speaking of geopolitics he tells you “I met a gentleman a Serbia that …”, a bit like those who quote the partisan grandfather. At one point the discussion ended up on NATO, and I told Carlo Sforzagreat liberal politician who managed to bring Italy into the BORN (When, according to the anti -American and communist narrative, it is believed that there has been “imposed” at the end of the war by the Americans), and on which three years of contemporary history at the University.
What does this have to do with the AI? It has to do with him because in the meantime Ballarani asked Chatgpt, but Chatgpt by Carlo Sforza knew more or less what is found in the discussions on the net or on Wikipedia. None of all the thousands of pages of essays and studies that are in texts under copyright. Hence the reason that brings those who develop the Ai to want to violate the copyright rules.
However, returning to the theft of books, which for the AI are only “data”, it must be acknowledged that it came out of a cause against goal. If it hadn’t been there we would not have known it, while as well as openai as Google are asking Trump to remove the scuffle imposed by copyright. At this point I think they are asking only an official permit, unofficially they already do it. Also because we always go back to the same speech: Openai (like all other companies) checks chatgpt, but Who controls Openi? Who controls the controller? This is why I believe that nothing can be done. It was just an American question, perhaps yes, but when there is a digital war with China on AI the Copyright problem, in one way or another, it will be circumvented, the stakes are too high. What do I think? I don’t know, I only know it will be inevitable.
On the other hand, in the sixties, even during the space competition between USA And Urss For those who came first on the moon, the technical leader of the Apollo missions was Wernher von
Braunbrilliant German engineer who had only one small stain: it was the same one who had designed the V2 for i Nazislaunched against London. But it was better to forget about it.