zlacker

[return to "The New York Times is suing OpenAI and Microsoft for copyright infringement"]
1. desire+M6[view] [source] 2023-12-27 14:42:01
>>ssgodd+(OP)
This is just rent seeking from dying media instead of working on creating something new in my view.

AI indeed is reading and using material sa a source, but is deriving results based on that material. I think this should be allowed, but now it is a fight who has better paid politicians pretty much.

I am open to hear other thoughts.

◧◩
2. ethanb+M9[view] [source] 2023-12-27 14:57:53
>>desire+M6
Here’s another thought: It’s good that there are real incentives to produce original content. Especially investigative journalism which is an extremely tough business financially — even without LLMs — but with lots of social value.

It would be silly to totally destroy the incentive to produce new technologies like LLMs, but so wouldn’t it be silly to destroy the incentive to produce original, high-quality content either for human or LLM consumption.

FWIW the LLMs are obviously the ones rent-seeking here, if you’re trying to use the term for its actual meaning instead of just “charge a subscription for something I don’t want to pay for.”

◧◩◪
3. Zpalmt+id[view] [source] 2023-12-27 15:15:53
>>ethanb+M9
How are the LLMs rent seeking? they are clearly providing value that people want to pay for..
◧◩◪◨
4. ethanb+we[view] [source] 2023-12-27 15:23:30
>>Zpalmt+id
This has to be one of the most abused terms on this website.

“People are willing to pay for it” is not even relevant to the question of whether it’s rent-seeking. Rent-seeking has to do with capturing unearned wealth, i.e. taking someone else’s work and profiting from it.

There is some portion of OAI’s (et al.) value that they themselves produce. There is another portion that is totally derivative of the data — other people’s work — they have trained on for free. A simple thought experiment can tell you to what degree OAI et al are “rent-seekers.”

Imagine a world where they had to enter into mutual agreements in order to train on that data. How much would the AI companies be worth? Not quite zero, but fairly close (Andreessen pretty much stated this IIRC). How much would the data producers be worth? The exact same amount or more.

[go to top]