General Discussion

That's complete nonsense. They stole all that IP because it's the only reason their AI tools work at all well. OpenAI admitted in court documents years ago that its AI would not work properly if limited to what's in the public domain and what they had a legal right to use.

Sometimes they try to confuse people by referring to what's "publicly available" - wording you're echoing - and hoping people will confuse that with what's in the public domain when it isn't.

And they're continuing to scrape websites every day, stealing more and more copyrighted IP. Even what's published under a Creative Commons license requires proper attribution and links, which AI slop rarely if ever provides.

It's theft. Period. Your echoing AI industry propaganda doesn't change that fact.

As for errors genAI makes - I've never claimed it gets everything wrong. But it can make mistakes at any time in any place in its results, and that's why even those robber baron AI bros include warnings about that, telling people to check its results. They're irresponsible enough to peddle flawed AI tools, but they want to shift legal liability for their own products' mistakes to the users.



I've read a lot by AI fans and addicts as well.

For instance, people using GPT-5.2, whose comments on its unreliability I posted on DU yesterday:

https://www.democraticunderground.com/100221158692#post3

I'll trust computer security firms' statements on AI code being a major security risk more than I'll trust someone using it and finding it a timesaver.

It's been well known since genAI tools were first released that it can be difficult to catch all their errors, and many users don't even try.

GenAI is inherently flawed, with no indication that will ever change - the newer allegedly "reasoning" models hallucinate even more than older models - and it's inherently unethical.