zlacker

If the future of humanity rests on access to old NYT articles, we’re fucked. Why can’t OpenAI try to get a license if the NYT archives are so important to them?

>>peyton+(OP)
They're not. They can skip the entirety of the NYT archives and not much of value will be lost. The issue is with every copycat lawsuit that sues every AI company out of existence. It's a chilling effect on AI development. Old entrenched companies trying to prohibit new ways of learning and sharing information for the sake of their profit.

>>solard+q2
Why don’t they train their AI on non-copyrighted material? It’s only fair for the copyright owners to want a share of the pie. I’d want one as well for my work.

>>lacrim+U6
Because they wouldn't have enough good quality training data then probably.

>>lacrim+U6
>It’s only fair for the copyright owners to want a share of the pie.

No it's not, it's pure greed. Everyone'd think it absurd if copyright holders dared to demand that any human who reads their publicly available text has to pay them a fee, but just because OpenAI are training a brain made of silicon instead of a brain made of carbon all the rent-seekers come out to try to take advantage.

>>lacrim+U6
Because no one forced them to, and the copyrighted dataset is much larger? It's like trying to teach your kids using only non copyrighted textbooks. There's not much out there.

Copyright is an ancient system that is a poor legal framework for the modern world, IMO. I don't think it should exist at all. Of course as a rightsholder you are free to disagree.

If we can learn and recite information, and a robot can too, then we should have the same rules.

It's not like ChatGPT is going around writing its own copycat articles and publishing them in newsstands. If it's good at memorizing and regurgitating NYT articles on request, so what? Google can do that too, and so can a human who spends time memorizing them. That's not its intent or usefulness. What's amazing is that it can combine that with other information and synthesize novel analysis.

The NYT is desperate (understandably). Journalism is a hard hard field with no money. But I'd much rather lose them than OpenAI. Of course copyright law isn't up to me, but if it were, I'd dissolve it altogether.

>>logicc+Xx
You know the NYT has to fork out money to build the content right ?

>>bamboo+651
Do you really think they're losing subscribers to ChatGPT...? Is there a single real person that thinks, "Oh, I don't need to pay the NYT anymore, I can just wait for the next OpenAI update six months from now and it'll summarize all the news for me"?

>>solard+YU
Ok, your reasoning escapes me. NYT has the right to sue and like any other business it’s holding onto their moat. Why would they let OpenAI train on their propery? Why wouldn’t they train their own AI on their own data?

Open AI is a business. NYT is a business. MS is a business. Neither will be happy when some other party takes something away from them without paying.

>>snowst+wr
Too bad. Quality costs. Share the profits with everyone then and nobody would be unhappy

>>solard+Q81
It's beside the point, the point is, money and time were spent producing that work, so why should OpenAI just be allowed to take it and profit from it, without at least attribution? It's absolutely ridiculous.

I saw an article the other day where they banned ByteDance's account for using their product to build their own, can you see the absolutely massive hypocrisy here?

It's fine for OpenAI to steal work, but if someone wants to steal theirs, it's not? I cannot believe people even try defend this shit. It's wack.

>>logicc+Xx
> No it's not, it's pure greed.

And Altman (Mr. Worldcoin) and fucking Microsoft are what, some gracious angels building chatbots for the betterment of humanity? How is them stealing as much content as they can get away with not greedy, exactly?