If they were, say, a charity doing this for the good of mankind, I’d have more sympathy. Shame they never were.
The best party about all this is that Disney initially took off by… making use of public domain works. Copyright used to last 14 years. You’d be able to create derivative works of most the art in your life at some point. Now you’re never allowed to. And more often than not, not to grant a monopoly to the “author”, but to the corporation that hired them. The correct analysis shouldn’t be OpenAI vs. Intercept or Disney of whomever. You’re just choosing kings at that point.
People do get sued for making songs that are too similar to previously made songs. One defence available is that they've never heard it themselves before.
If you want to treat AI like humans then if AI output is similar enough to copyrighted material it should get sued. Then you try to prove that it didn't ingest the original version somehow.
A legal argument would be needed to argue the other way. This argument would imply granting LLMs some degree of human rights, which the very industry profiting from these copyright violations will never let happen for obvious reasons.
Also, it is only a matter of time until one of those employees (thanks to free will and agency) will whistleblow, it doesn’t scale, etc.
Frankly, the fact that such a big segment of HN crowd unthinkingly buys big tech’s double standard (LLMs are human when copyright is concerned, but not human in every other sense) makes me ashamed of the industry.
You would be except for the fact that publishing stuff on the web gives people an implicit license to download it for the purposes of viewing it.
OpenAI (and Google and everyone else) is creating a publicly-accessible system that produces output that could be derived from copyrighted material.
That‘s confidently and completely wrong.
I agree, but the original author might get butthurt if you distribute it. Realistically copyright law in the US is a mess when it comes to weird pieces of art.
It is disingenuous to imply the scale of someone buying books and reading them (for which the publisher and author are compensated) or borrowing them from the library and reading them (again, for which the publisher and author are compensated) is the same as the wholesale copying without permission or payment of anything not behind a pay wall on the Internet.