If Kasparov uses chess programs to be better at chess maybe we can use copilot to be better developers?
Also, anyone, either a person or a machine, is welcome to learn from the code I wrote, actually that is how I learnt how to code, so why would I stop others from doing the same?.
Judging by the majority opinion in this thread, it seems pretty clear GitHub could have asked and gotten enough people to opt-in to have no problem training their model. They probably would have been thrilled to do it and proud of being included in the training data.
But the preference of the majority does not override the conditions placed by people who prefer not to participate.
No human perfectly reproduces the learning material they used.
If that was true, one might as well just higher engineers from Twitter and make a new platform from the code they remember!
Well, we humans do it occasionally. You probably remember a few specific code snippets in your lang of choice because they kept annoying you/you love them/you wrote them a lot. So if I would put you in the exactly right situation, you would indeed reproduce code verbatim.
So does Copilot.
I am not trying to insinuate that Copilot works like a human, but it is literally the same situation.
If Kasparov uses chess programs to be better at chess maybe we can use copilot to be better developers?
Also, anyone, either a person or a machine, is welcome to learn from the code I wrote, actually that is how I learnt how to code, so why would I stop others from doing the same?.