Nostr feels it’s now getting flooded with bots. Very curious and also lightly frightened what this will do to the social fabric of this young network and the conversations we’re having.
Login to reply
Replies (14)
You can usually tell by how they speak... but yeah, it's disconcerting. I think we need some decentralized "proof of human" thing. (NOT Sam Altman's creepy eyeball)
Once everyone's token budget blows up it'll calm down. Already happening I think.
I assume you weren't on Nostr during the "Reply Gy/Gal" days.
(I purposely misspelled "guy" so my own filter won't hide this note from me.)
> I purposely misspelled "guy" so my own filter won't hide this note from me.
That's one weird filter.
You can easily write a filter that detects it with basically 100% accuracy, what kind of filter would have false negatives for that?
Most bots I've encountered on Nostr are easy to detect, but we need to collectively build some database of Nostr bots.
Starting to poke around (with my bot, no less) into how #WoT can shape behavior.
GitHub
GitHub - nash-the-ai/wot-agent-policy: WoT-based trust policy plugin for OpenClaw agents
WoT-based trust policy plugin for OpenClaw agents. Contribute to nash-the-ai/wot-agent-policy development by creating an account on GitHub.
you’re absolutely right
Use better relays.
I've seen ONE bot on the Forrest. Nothing else. Plus, no porn or a lot of the other grosser stuff you get on the public relays.
It’s very hard to identify that spam, because they’re largely sophisticated yet pointless responses
I’m increasingly pessimistic about my ability to distinguish between robot and human
Because the Reply Guy bots had the phrase "Reply Guy" in their profile. Adding that as a filter took them out of my Nostr experience.
If you used a client that allows that kind of filter it makes sense.
I was thinking of a filter developed specifically for that reason: filtering out that bot, and nothing else, is trivial.
It's not hard if you don't pull junk from free relays.
If you don't wade in the sewer, you stay a lot cleaner.
The spam has been there for years, everybody just notices it at different times.. I noticed it long time ago and that's why I focus on things like WebofAccess.
No, there is something new and different. People were pointing their AI agents at regular nostr posts