2 authors say OpenAI 'ingested' their books to train ChatGPT. Now they're suing, and a 'wave' of similar court cases may follow.

Researchers pay for publication, and then the publisher doesn't pay for peer review, then charges the reader to read research that they basically just slapped on a website.

It's the publisher middlemen that need to be ousted from academia, the researchers don't get a dime.

It's generally not the creator who gets the money.

There’s also GPL, which states that derivations of GPL code can only be used in GPL software. GPL also states that GPL software must also be open source.

ChatGPT is likely trained on GPL code. Does that mean all code ChatGPT generates is GPL?

I wouldn’t be surprised if there would be an update to GPL that makes it clear that any machine learning model trained on GPL code must also be GPL.

Closed source software isn't going to be available to AI scrapers, so this only really affects open source projects and open data, exactly the sort of people who should have more protection.

The point of open source is contributing to the crater all of humanity. If open source contributes to an AI which can program, and that programming AI leads to increased productivity and ability in the general economy then open source has served its purpose, and people will likely continue to contribute to it.

Creative of Commons applies to when you redistribute code. (In the ideal case) AI does not redistribute code, it learns from it.

And the increased ability to program by the average person will allow programmers to be more productive and as a result allow more things to be open source and more things to be programmed in general. We will all benefit, and that is what open source is for.