4 ms·

I've been trained on open source code, and there are likely many algorithms that I've internalized that are very similar to the "standard" way of performing an

I've been trained on open source code, and there are likely many algorithms that I've internalized that are very similar to the "standard" way of performing an operation. Is there a reason why an AI being trained on the same open source code isn't a similar situation? I agree that wholesale pasting of code chunks is an issue, but that hasn't been my experience with Copilot. I'm not arguing for Copilot here...I'm genuinely curious why this would be considered any different.

layer8 4y ago
Humans can reasonably distinguish between when they are plagiarizing and when they are just applying their experience and knowledge. Copilot presumably isn’t able to make that distinction, and, arguably, so aren’t the consumers of Copilot’s output.
bakugo 4y ago
>Is there a reason why an AI being trained on the same open source code isn't a similar situation? You are a human. You know what's right or wrong. You know you can't just copy code 1:1 from public repositories without respecting their license. The AI doesn't know and doesn't care. It's a common problem with creative AIs that they will occasionally regurgitate near 1:1 copies of their training data, and I don't think it's an easy problem to solve. >I agree that wholesale pasting of code chunks is an issue, but that hasn't been my experience with Copilot. The article provides several examples of it happening. Just because it hasn't regularly happened to you doesn't mean it doesn't happen.
- mdswanson 4y ago
  I don't deny that there are examples that appear to be wholesale copying, and that is definitely an issue to be addressed. No doubt. What I don't understand is why the rest of the service (where it doesn't appear to be pasting existing code) is being maligned when it behaves like a more powerful version of autocomplete.

4 ms·

I've been trained on open source code, and there are likely many algorithms that I've internalized that are very similar to the "standard" way of performing an

by mdswanson 4y ago

layer8 4y ago
Humans can reasonably distinguish between when they are plagiarizing and when they are just applying their experience and knowledge. Copilot presumably isn’t able to make that distinction, and, arguably, so aren’t the consumers of Copilot’s output.
bakugo 4y ago
>Is there a reason why an AI being trained on the same open source code isn't a similar situation? You are a human. You know what's right or wrong. You know you can't just copy code 1:1 from public repositories without respecting their license. The AI doesn't know and doesn't care. It's a common problem with creative AIs that they will occasionally regurgitate near 1:1 copies of their training data, and I don't think it's an easy problem to solve. >I agree that wholesale pasting of code chunks is an issue, but that hasn't been my experience with Copilot. The article provides several examples of it happening. Just because it hasn't regularly happened to you doesn't mean it doesn't happen.
- mdswanson 4y ago
  I don't deny that there are examples that appear to be wholesale copying, and that is definitely an issue to be addressed. No doubt. What I don't understand is why the rest of the service (where it doesn't appear to be pasting existing code) is being maligned when it behaves like a more powerful version of autocomplete.