OpenAI releases a 15-page set of policy suggestions for the White House, Urging for reduced regulation; Calls for a strategy to preserve American AI models ability to learn from copyrighted material.
You're viewing a single thread.
When Aaron Schwartz was in trouble for downloading copyrighted articles everyone was on the side of "copyright laws are dumb and need to be changed".
Now it's more popular to hate on AI and so now people want to see strict adherence to copyright law.
It Mmkes it seem like people lack real convictions on the issue and are just being led around by memes.
Copyright law is terribly implemented and needs to change. This isn't new and doesn't become less true because your favorite memes want you to dunk on AI.
Copyright laws are dumb and need to be changed.
And I also think that the laws that currently exist should apply to openAI. I don't see any contradiction there.
The current system where regular people can get screwed over for torrenting movies but techno-oligarch wannabes are free to ignore the law is the absolute worst of both worlds.
To be clear, this isn't a discussion about removing copyright laws. This is a discussion about specifically big data collecting tech companies being immune to the laws which still apply to everyone else.
I never suggested or implied that copyright laws need to be removed.
It does appear that OpenAI's position is "copyright laws are dumb and need to be changed" which, during the Aaron Schwartz story, was the position of the community.
Now, since the entity involved is an AI company, we're seeing people who're on the side of using copyright laws to punish infringers because they don't like AI.
Either copyright laws need to be changed or they don't need to be changed. Someone's position on the topic shouldn't change based on who is being negatively effected by said laws.
People are being cynical about the laws applying equally to Big Corps vs regular people.
If we make a special case for abolishing copyright if it means you're training an AI model, does that mean that now everyone can download copyrighted material if they do some form of locally hosted training?
The answer will probably end up being: one rule for the corporations and another for individuals.