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
Conversation
Github's not denying that copilot output is derivative of the training data; instead, they're claiming fair use (so long as nontrivial amounts of code aren't reproduced exactly). HN post from Github's current CEO: news.ycombinator.com/item?id=276783
(not sure at all about this myself...)
2
4
This Tweet was deleted by the Tweet author. Learn more
That's not an accurate portrayal of what that case what about or why Google won the case.
11500 lines of class declarations defining public APIs is not at all the same thing as implementation code, as was determined by the court.