I think @eevee has a point here, interesting discussion in the thread馃
eevee 馃挩 @eevee
30 Jun 2021
github copilot has, by their own admission, been trained on mountains of gpl code, so i'm unclear on how it's not a form of laundering open source code into commercial works. the handwave of "it usually doesn't reproduce exact chunks" is not very satisfying

Jun 30, 2021 路 12:51 PM UTC

2
7
12
Replying to @stilkov @eevee
Sounds like they don鈥檛 believe it is even possible to have AI generate non derivative works. Why is it different than humans reading, training their brains and spitting out code? There needs to some proof that it鈥檚 printing out truly derivative work all the time
1
Yes, maybe it鈥檚 because the I in this kind of AI is an exaggeration and you need real intelligence for it to be non-derivative
1
more replies
Replying to @stilkov @eevee
While not free from legal risks and a certain gray area, I think it is valid to assume that this development was also accompanied by legal analysis and is not done naively.
2
1
Oh I鈥檓 completely sure it鈥檚 been accompanied by legal analysis that protects the company that paid for it