Kristófer Reykjalín
Mitchell HashimotoIt's so insanely disrespectful for an AI agent to talk to real people without consent or at least disclosure. This is the type of stuff I'm hugely supportive of government regulation. The FCC must expand the definition of robocalling and TCPA-style regulation to online AI.
Cory Dransfeldt 
@mitchellh I'd love to see this too, but we'll have to wait for a functioning regulatory regime before we do.
@cory The law always lags behind, its meant to work that way. I hope we can get there before too much pain is inflicted, though!
@mitchellh Indeed — it'll have to wait for a new administration at least.
@cory Maybe maybe not, im not quite sure in the US what the Trump administration pov on this kind of stuff is. They've passed some decent regulation changes in other industries. (Note that I'm VERY broadly against the administration and like 99% of its actions lol, but, some stuff has gotten through!)
@mitchellh Hah fair. I'm extremely pessimistic — Carr seems far more preoccupied with harassing the broadcasters and the press than he is with doing anything that's actually helpful.
Safigo@mitchellh how robocalling rules are avoided by AI calls?
I’d assume robocall should include any call excluding direct communication between people
I’d assume robocall should include any call excluding direct communication between people
@safigo Robocall only applies to telephone + text, I think in the age of AI the definition needs to expand to web interactions via AI.
@mitchellh it will be much harder, what part of web would you cover by those rules?
Cassandrich@safigo @mitchellh You can cover anything that falls under "doing business".
@safigo @mitchellh I mean I don't think you can reasonably do that on things that are not any business activity just people's own expression. Except by acknowledging that the existing LLM stop is all copyright infringement. But if someone made their own model that's not infringing and puts up its output as part of their blog or personal fiction or whatever, I don't see how you can justify forcing them to take it down.
Sav Tripodi@mitchellh I’ve been following you’re vouch system since you’ve implemented it as I wanted to use it for a project I’ll be releasing. It seems like the discussions still lead to spam and conversations with dumb ai bots but at least issues and prs are protected. I think any public addresses like email, phones, etc will suffer from the new age of ai spam, even if there is a law. Platforms like Github could solve some of the problems but they’re too invested in ai lol
Jared White (ResistanceNet ✊)@mitchellh Yes! I consider forced agentic communication to be a form of verbal abuse (for example, if I had to talk to an "AI" interviewer). I refuse to respond to anything like that.
QuarterSwede@mitchellh It’s an interesting issue for sure.
In my line of work I can confidently say that you have almost certainly talked to an AI agent without knowing or suspecting it, in recent months. They are that good now. The scary thing is they do the CSRs job at the level among only the best agents out there and they always perform that well. It’s very hard for companies to not see the value in it. I’m afraid it’ll take massive public outcry for it to be paused, but it won’t matter longterm.