Does it work though? The big LLM crawlers do not read llms.txt so will they read and follow the same instructions as HTML?

Where did you see the big crawlers don't read it? Anthropic does.. they're pretty big.

If I look at the logs of sites that Anthropic crawls (and in some cases references in answers) and there are no requests for llms.txt

Someone has to have done or is doing an experiment with this right? I also think that if it was an actual profitable thing then we would know about it pretty quickly. It would pop up everywhere

Apparently new checks in Chrome Lighthouse are checking for the existence of the file.

https://searchengineland.com/google-llms-txt-chrome-lighthou...

Yeah I want to know how many donations they get

I have no idea, in theory it might catch some miss-configured agents off-guard

This is not for LLM data crawlers, it's for LLM agents the end users are running on their machines