Hacker News Reader

Comment by blondie9x | Hacker News Reader

Of course it did. This is how AI is trained and becomes profitable with current business models. It uses your content without paying you for it to generate new content.

AI can be one of the most powerful forces in the world but it needs to pay content creators. If people stop creating new content for AI to train on then it will get stale.

AI can be the product of our dreams and our passions but we need to make sure those who choose to create and share content to train it are treated fairly and compensated when applicable.

View on HN

This isn't a copyright lawsuit. Those lawsuits are already in progress.

This lawsuit is alleging that OpenAI trained GPT-3 and -4 on inadvertently published information. Web crawlers are very good at finding things you wouldn't expect to be public; there's techniques you can use to, say, abuse Google to search for such things.

When companies stop making “document1.pdf” publicly facing, while expecting “document2.pdf” to remain confidential, then we can talk.

I'm fine with talking to them now - I don't personally find your argument very compelling; there is almost always a step further that someone wont consider that someone else will think is obvious.

So if i enumerate a host and find a git dump that includes private api keys i can do whatever i want with them?

Accessing a system you don't have permission to access because it's misconfigured is still illegal lmao. People have gone to jail for doing exactly this.

AI should be a tool for creators and the benefits come from its use. Instead of trying to own ideas, creators should only protect the unique way these ideas are expressed, that's what copyright should cover.

I think that’s the same thing to be honest. You are more talking about protecting the artistic process rather than output. But the output is what is used to train. Not the process.

What an entitled attitude.

Reply on news.ycombinator.com