Hacker News Reader

You really only need one example to offset this anecdote, so here you are: https://mobile.twitter.com/mitsuhiko/status/1410886329924194...

Copyright violations are a genuine concern from the outputted code, GitHub themselves have admitted it may emit raw training data rarely.

View on HN

There is logic to ensure that copilot does not emit exact duplicates of code in the training set... but that logic is significantly newer than that tweet.

Link? I couldn't find anything "significantly newer" than 7/2/21 (though I'm sure GitHub is doing a lot here). They had this blog post 6/30/21 regarding efforts on avoiding raw code: https://github.blog/2021-06-30-github-copilot-research-recit.... They concluded:

> We will both continue to work on decreasing rates of recitation, as well as making its detection more precise.

Source: I work on the copilot team.

Was that decision informed by legal or product? Because derivative works are still derivitative works even if you don't replicate the original verbatim.

I mean, it was informed by both, but basically everyone thinks it's a good idea.

Reply on news.ycombinator.com