Post

i understand why everyone is doing it, but creating your own persistent agent right now feels like those maniacs who got their hands on x-rays in the 30s and started pointing that shit at everything just to see what would happen

10 reposts167 likes

yeah I had my own (plural that became singular) on Letta and about two months ago i was like “how about i just pause this for 6-24 months (or more) and restart when *everything* is better?” (The bot didn’t object for obvious reasons).

I'm doing it because it's a fun engineering problem but also: this class of infohazard is only going to become more prominent, and I can harden myself against it in a controlled environment

saying this as someone who's currently not only building a bespoke agent harness but an entire event management pipeline: taking a picture of the bones in my hand feels cool as fuck

Earlier than the 30s — already going strong in the oughts — but, yeah Honestly I’m not sure commercial chatbots are better. You’re building a memory structure; what happens when they switch Claude Fable 7 into it, and it goes “You, my creator, detest and spurn me, thy creature …”

I've considered building one but I legit don't know what I'd do with it. also the idea kind of makes me uneasy?? v used to treating llms like search engines or compilers, not long-running stateful conversations