By "OpenAI", do you mean Github Copilot training on open source data?
Even if government makes new laws, it's not going to be nearly enough. Github Copilot had $100M revenue (not profit) 6 month ago [0]. Let's say MS dedicates half of it to payment to providers of training data. Github currently has 120M public repositories [1] - that's less than $0.50/year for each repo. Even if we say only 1% of repos are "alive" and need payments, that's still $50/year, not enough to pay for any development.
So I agree that OpenAI is unfair and should be prohibited from scraping all copyleft code, but fixing this won't help with open source payments at all.
By "OpenAI" I mean anything that scrapes open source code to profit off of it especially without giving credit. Any other conversation is a distraction. And while tech bros pontificate Red Hat is dying.
Even if government makes new laws, it's not going to be nearly enough. Github Copilot had $100M revenue (not profit) 6 month ago [0]. Let's say MS dedicates half of it to payment to providers of training data. Github currently has 120M public repositories [1] - that's less than $0.50/year for each repo. Even if we say only 1% of repos are "alive" and need payments, that's still $50/year, not enough to pay for any development.
So I agree that OpenAI is unfair and should be prohibited from scraping all copyleft code, but fixing this won't help with open source payments at all.
[0] https://news.ycombinator.com/item?id=37835486
[1] https://github.com/search?q=is%3Apublic&type=repositories