it's theft. There's been plenty of evidence of that, from internal documents at different companies to a former Google CEO advising that AI startups should just steal IP and bet that they won't get sued unless their company is successful, and if they are successful they can afford enough lawyers to fight the lawsuits. I posted an LBN thread about that CEO. I've. been explaining this to people here for years.
And I realize most people who like AI would prefer avoiding the fact that these are unethical, illegally trained tools based on theft. I've had to point out to AI-using academics on X - academics worrying that AI tools they like to use will steal their work, and their students' work, for training - that they're being both foolish because they can't trust the AI companies not to steal it, and hypocritical because the only reason those tools work at all is because so much intellectual property was stolen before the tools were released. The value of those AI tools is almost 100% in the stolen IP, which the AI bros have no intention of paying for. OpenAI has admitted in court documents that they have no hope of ever making a profit if they have to pay for what they stole.
https://www.tomshardware.com/news/ai-doesnt-learn-like-people-do
= new reply since forum marked as read