[elondaits]

It might be a problem to treat it as copyright. Copyright applies to reproduction, distribution, public performance... if I go to a library or bookstore and I read books and look at their covers, copyright does not apply. Would an android that walks around learning things be subject to copyright? To what extent does it need a body and mobility to be more like a person and less as a scraper?

It might seem stupid, but I worry that if copyright begins applying to "mining" then the next thing is that it applies to humans watching things.

Of course, if an AI re-creates copyrighted content, copyright should apply. Just like it applies when I redraw and sell the Mona Lisa, but not when I store it in my memory. I would pass on the responsibility to users. I don't fear my use of Github Copilot because it's far from infringing any reasonable copyright... then again, I'm assuming the most likely way to infringe copyright with GPT is to use a prompt that almost explicitly requests it.

[chrismorgan]

Funny thing is that recreations are not necessarily covered by copyright law. The clearest example of this is fonts, where (using imprecise terminology but I think it’ll be clear enough) the US only grants copyright protection to font files, but not the shapes—so tracing a commercial font is perfectly legal, even if you happened to end up with an identical result (though good luck proving to a court that that’s what you did).

[rockemsockem]

I mean, today it's pretty easy to prove that if you're setting out to copy one of these fonts. Just film the whole process and upload it to YouTube.