LLM powered chatbots are pretty big, these days. There’s constant speculation about the big LLM companies putting these chatbots into devices that’ll just let you talk out loud to them, and they’ll chat back to you and do your bidding. Recreating “Her”, as Sam Altman unsubtly yearns to do.

I hate it.

Putting aside any qualms one might have about the underlying technology, of which I have a few, the interface paradigm is unbearable to me. I cannot fathom wanting to have a verbal back and forth with my computer. I particularly can’t see it being something I’d ever want to do in public, in an open office or a coffee shop or whatever.

I’d think this might be a programmer thing, craving specificity and precise control over what’s happening, but “vibe coding” is a notable part of this trend so that’s presumably not it.

The appeal of “AI companions” also doesn’t register to me. If we had AI, and not spicy autocomplete, I could understand it… but the fundamental knowledge that you’re talking to a predictive model ruins any appeal I could see.

Am I so out of touch? No. It’s the children who are wrong.

(That said, I’m amused that my previous use of the “opinion” tag on this site turns out to have been me complaining about a similar issue with disliking everything switching from text to video.)