- cross-posted to:
- aicopyright@lemm.ee
- cross-posted to:
- aicopyright@lemm.ee
‘Impossible’ to create AI tools like ChatGPT without copyrighted material, OpenAI says::Pressure grows on artificial intelligence firms over the content used to train their products
If OpenAI is right (I think they are) one of two things need to happen.
For number 1. Good luck for all the reasons we all know. Capitalism must continue to operate.
For number 1. Good luck because those in power are mostly there off the backs of those before them (see Disney, Apple, Microsoft, etc)
Anyways, fun to watch play out.
There’s a third solution you’re overlooking.
3: OpenAI (or other) wins a judgment that AI content is not inherently a violation of copyright regardless of materials it is trained upon.
It’s not really about the AI content being a violation or not though is it. It’s more about a corporation using copyrighted content without permission to make their product better.
If it’s not a violation of copyright then this is a non-issue. You don’t need permission to read books.
AI does not “read books” and it’s completely disingenuous to compare them to humans that way.
That’s certainly an opinion you have
Backed by technical facts.
AIs fundamentally process information differently than humans. That’s not up for debate.
Yes this is an argument in my favor, you just don’t understand AI/LLMs enough to know why.
I could say the same about you, considering I’ve watched you peddle false information for months about this subject.
AI learns differently than humans. That isn’t a fact up for debate. That’s one of the few objective truths around this industry.
Similarly I don’t read “War and Peace” and then use that to go and write “Peace and War”
There’s no open source without copyright, only public domain
If all is public domain, all is open source
Open source also includes viral licenses like the GPL. Without copyright, the GPL is not enforceable.
It doesn’t have to be. One leak and the code is open for all