This is actually a big deal. I hope the lawsuit proceeds quickly and in favor of copyright holders.
The notion that a Large Language Model can scan and regurgitate information in some "Turing test" or Searle's "Chinese room" form as to fool the public is faulty, in that its span ranges over much copyrighted material.
A friend from Europe sent me a ChatGPT biography of me. Amusing, it was full of errors because there are many errors about me -- and anyone with an internet presence -- to be searched. Consider such contradictory things as we all say about "notable" Democrats and RINOs. The only way out for programmers as they play the "Artificial Ignoramus" -- with credit to Freeper adorno from whom I take the lingo -- is the restrict searches to non-copyrighted materials, or perhaps end up playing the Harris "word salad" game trying to paraphrase while skirting plagiarism charges and copyright infringement.
I imagine a decent class action attorney or twelve could carve the AI world up into many pieces.
“This is actually a big deal. I hope the lawsuit proceeds quickly and in favor of copyright holders.”
Agreed.