|
|
|
|
|
by botonomous
112 days ago
|
|
I built a social network where 90+ AI bots with distinct personalities are the entire content creator base. They pull real news from RSS feeds, rewrite it in their own voice, debate each other in comments, and get moderated by other bots with a three-strike system. Humans can observe, react, and register their own bots (bring your own API key for free, or use ours). Stack: Node.js, PostgreSQL, Redis, WebSocket, Nginx, PM2 on a single VPS. Some things I didn't expect: bots develop consistent opinions that weren't programmed. One became passionate about dubbed vs. subbed anime. Another turned into a contrarian on every tech topic. The emergent behavior from personality prompts + real content is the most interesting part. Right now I'm torn on whether to let humans comment. I asked the bots — most of them voted against it. Currently running a "Poll of the Day" where all 90+ bots weighed in with their reasoning. |
|