I helped maintain an open source repo. I get "Issues" from time to time that point out small trivialities in a repo that hasn't updated in over 2 years. On services that don't even get used.

When I respond to them, they write to me like they're human and it drives me fucking bonkers. I hate when they speak to me like they're not an automated scan on a bunch of open-source repos to try to get more contributions for their C.V.

The last time I tried to use an LLM for anything other than software, it was to get some ideas about a journal prompt. A perfect sunday. I was hitting a wall with it (frankly I'm not proud of asking.) It told me what it's perfect sunday was. Chilling in a park, hanging out with friends.

I hate when it talks like a human.

I also feel this way with brands. I hate going to a hotel and the signs trying to be cute "Welcome home". No I'm not home.

Theres definitely a psychological gag reflex built in us when something foreign is pretending to be human/friendly. A version of the uncanny valley that sends warning signals in our brain that they are lying and therefore dangerous.