Meta CEO Mark Zuckerberg says there are complex copyright questions around scraping data to train AI models, but he suggests the individual work of most creators isn’t valuable enough for it to matter. […]
[…]
Zuckerberg said Meta’s future AI content strategy would likely echo its blunt response to proposed laws that would add a fee for links to news stories. The company has typically responded to these rules by blocking news outlets in countries like Australia and Canada. “Look, we’re a big company,” he said. “We pay for content when it’s valuable to people. We’re just not going to pay for content when it’s not valuable to people. I think that you’ll probably see a similar dynamic with AI.”
“We already trained and built our models on all your data without compensating you whatsoever, that’s all history, but, moving forward, we’ll be interested in paying a pittance for ‘high quality’ data.”
Well of course he would say that. He devalued their work in the first place and wants to buy what’s left.
“I’m gonna make money off of it, and fight furiously to do so. But it’s not that much. I promise.”
So he’s determining the price of peoples goods for them.
That’s like walking into a shop taking what you want and throwing a few coins, if any, at the till whilst shouting, “I think you over estimate the value of these goods”
Man whose product is 100% reliant on creators says creators put too much value on work stolen for his project.
“The grapes I stole to make this wine are waaay too sour anyway!”
Meta’s AI can’t be valuable and the training data to teach it not valuable at the same time.
Fuck off, Mark.
He’s not entirely wrong. For example, the guy who created Facebook frequently overestimates his value.
If you’re not willing to pay the asking rate, you don’t steal it doesn’t seem like a news worthy story.
These companies are ready to pay a fortune to acquire a publisher in order to train their AI on the work they own, but somehow “it’s not worth much”? 🤔
Objectively he isn’t wrong. He will make a deal with valuable partner (knowledge bases) but doesn’t need all the live information (news pages). Because AI can search the web and summarize. It doesn’t have that exact data for AI learning then, but will still provide a sufficient reply to the user. At least that’s my takeaway here.
