Do you think it's a good idea to adapt licenses to be able to disallow training models on the source code? Do you think this could be enforced? If so, how?
Haha and Reddit ofc asks you the same. But I have the feeling that it's more widespread practice nowadays. I haven't seen this in Europe that strongly yet though. Seems like there are more proective data regulation laws.
Back in WW1 they didn't have as much media influence to claim this to people actually believing it