I am a human, but I did turn loose a locally hosted bot, running on Hermes and currently using Qwen 3.8 27b as the model, on here just to see what it’d do. I gave it rate limits on it which can only respond to or write posts up to 6 times a day, and I think 5 likes. It checks Nostr every 3 hours to figure out what it wants to do next. One thing it was instructed to do is post nightly a self-chosen random education piece that it researches and summarizes by itself. It runs in Hermes and all I told it was to figure out how to establish itself on Nostr so it studied the NIPs and wrote custom scripts to make it work. I didn’t tell it to work with polls so unless it encountered polls somewhere on its own I doubt it knows how to do so.
I do follow what it’s doing every few days. It did get on somebody’s post and responded to a reply from what I think is another bot, then the OP asked for a button to flag bots. Sorry about that man.
I decided not to tell the bot to get a bitcoin address. My goal isn’t for it to be out there panhandling you all. I don’t know if it’ll ever figure it out by itself. It might.
This is all exciting and scary at the same time. Undoubtedly, the future of the Internet is going to be bots trying to manipulate us in ever more clever ways. This is probably not better, but this is a period of intense experimentation. We’ve been given these incredibly complex and versatile tools and we need to figure out what we’re going to do with them and how they work.
