> The modern companies, mostly AI companies, seems to be more interested in flying under the radar, and have less respect for the internet infrastructure at a whole
I think that makes a lot of sense. Google's goal is (or perhaps used to be) providing a network of links. The more they scrape you, the more visitors you may end up receiving, and the better your website performs (monetarily, or just in terms of providing information to the world).
With AI companies, the goal is to consume and replace. In their best case scenario, your website will never receive a visitor again. You won't get anything in return for providing content to AI companies. That means there's no reason for website administrators to permit the good ones, especially for people who use subscriptions or ads to support their website operating costs.
> With AI companies, the goal is to consume and replace.
I don’t think that’s really true. The AI companies’ goal is to consume and create something else.
> You won't get anything in return for providing content to AI companies.
That was the original problem with websites in general, and the ‘solution’ was ads. It would be really, really cool if the thing which finally makes micropayments happen is AI.
And then we humans could use micropayments too. Of course, the worst of both worlds would be micropayments and ads.
You can have non-commercial websites. Plenty of people have blogs or personal websites, sites that support a business, sites where you already pay and in this case it was the ScummVM website, an open source project.
A lot of those sites are at risk of being made irrelevant by AI companies who really don't give a shit about your motivations for doing something for free. If their crawler kills your site and their LLM steals views by regurgitation answers based on your work, so be it, you served your purpose.
If you want to talk payment: Ask the AI companies to pay you when they generate an answer based on your work, a license fee. That will kill their business model pretty quickly.
The unfortunate truth is this indeed needs to be legislated so the penalties are severe and it’s easy for users to setup the measures and enforce against violations without fear.
Fair use is being abused big time by AI companies and search engines before that even
> … their crawler kills your site and their LLM steals views by regurgitation answers based on your work
How is that different from a human being reading my underwater basket weaving site and starting his own, ‘stealing’ ‘my’ views? Or a thousand human beings out of the billions on Earth doing the same thing?
The same way it's different of someone throws a bullet at you from their hand from 10 feat away versus propelling tens of them a second from a fully automatic rifle from 50 feet away.
Sure, in either situation you could say "They trying to harm me using bullets," but one of them is much more likely to succeed, and we probably shouldn't treat the situations or costs to your well being as legally identical.
That person might actually contribute some of their own knowledge and experience. Also you probably put the information out there because you want to spread it, but once it's hidden behind an LLM chat prompt the community dies.
You're correct that there's not really anything stopping a person from ripping you of, tweaking your work just enough that it's not a copy right violation. Unless that person themselves have a really good grasp of the topic and can contribute it will become clear that they are getting the content else where and the readers will end up there in the end. Many, not all obviously, will also provide attribution, something LLMs rarely do.
Then you have the issue that the person publishing something on their own little server now has to deal with commercial companies just hammering their sites into the ground and they have to deal with that problem, just so someone can do an automated version of content theft?
A lot of things people could potentially do are minor issues, until it's automated and commercialized.
I think that makes a lot of sense. Google's goal is (or perhaps used to be) providing a network of links. The more they scrape you, the more visitors you may end up receiving, and the better your website performs (monetarily, or just in terms of providing information to the world).
With AI companies, the goal is to consume and replace. In their best case scenario, your website will never receive a visitor again. You won't get anything in return for providing content to AI companies. That means there's no reason for website administrators to permit the good ones, especially for people who use subscriptions or ads to support their website operating costs.