[eptcyka]

Humans don't learn an algorithm by memorizing a particular implementation character by character.

[ignoramous]

That's all the more reason for the utility of solutions like Copilot? Humans are limited in both time and memory.

Though, GitHub would do well to also bake-in approp attributions if a significant portion of the generated code is a copypasta.

[remix2000]

Neither does copilot.

[eptcyka]

But it does though. There have been many times where this was the case.

[Kiro]

It only happens if you bait it really hard and push it into a corner. That's not representative at all. I use Copilot to write highly niched code that's based on my own repo. It's simply amazing at understanding the context and suggest things I was about to write anyway. Nothing it produces is just copypasted character by character. Not even close.

[bamboozled]

As others have pointed out, it means the model contains copyrighted material. So I guess that’s totally illegal. Like if I ripped a Windows ISO, zipped it up and shared it with half the world. You know what would happen to me don’t you ?

[Kiro]

Not the same thing at all. The data isn't just sitting there in a store inside the model that you can query. No-one would be able to look at the raw data and find any copyrighted material, even if all it was trained on was copyrighted code (which I agree is an issue).

[ChatGTP]

There’s a lot of misconceptions here but LLMs and stable diffusion have spat out copyrighted material verbatim.

So that’s not accurate.

[golergka]

And airplanes don't flap their wings, but we still agree that they're flying, just as birds do.

[vadiml]

There are people who do it... I personally know a guy whit photographic memory

[j16sdiz]

He don't get an exemption from copyright law, or do he?