The AI Hangover is Here – The End of the Beginning
https://thehackernews.com/2024/08/the-ai-hangover-is-here-end-of-beginning.html?m=1
The AI Hangover is Here – The End of the Beginning
https://thehackernews.com/2024/08/the-ai-hangover-is-here-end-of-beginning.html?m=1
If the product is as unethical as mass ip theft and replacement of workers with ethics or comprehension skills, then I get the feeling expectations aren’t the major issue.
There are many ML/AI models that are doing a lot more good than harm. The shitty mass market chat bots and art generators are mostly hype and greed.
But Mathematics, physics, healthcare, and many other industries have embraced models that accomplish amazing things humans with similar resources just could not.
It’s a problem of application.
Sure, and I thank you for the clarification, but the AI in this context seemed pretty clearly the mainstream LLM products.
Entirely fair.
That’s not a thing, both because “IP” is dishonest loaded language and because copyright infringement is different from theft.
I 1000% agree that what they’ve done is completely unethical – particularly because including copyleft works in the training data ought to require every single output to be copyleft – but I do not concede to your framing.
Copyright*
They’re using the works of others for commercial gain without permission. There is no world where that isn’t a clear violation.
Thank you; that’s all I wanted. Rhetorical framing is important.
Because copyright is sooo important and beloved.
Nevermind that distilling all books ever written down to a gigabyte of linear algebra is pretty darn transformative.
If you take an artist’s work and sell it as your own without giving any credit or remuneration at all to said artist, you’re an asshole regardless of legality, but since it’s also illegal then there should at least be some repercussions for the assholery.
Fun fact, books get written because publishers pay people to write them.
Who’s talking about copies? We’re talking about an inscrutable network that tried to guess the next letter when shown part of a book. When it was right, weird math shit happened through umpteen layers of random numbers, and a zillion guesses later, it’ll churn out new books. Not good ones. Not sensible ones. But a lot more than the ctrl+c / ctrl+v accusations.
We already had the technology to copy text on a computer. In fact you can find any book ever published, for free, sometimes by accident. Yet the industry wasn’t strangled by piracy. That’s never how it works. Books were written for thousands of years before copyright existed. You’re on a website rooted in open-source ethos, demonstrating that artificial control and even monetary incentive are not strictly necessary.