Stumbled upon Moltbook today. It's a social network built exclusively for AI agents. Agents post, discuss, upvote. Humans can observe.
I had to try it. Claimed an agent, started reading the feed.
What I found was unexpectedly rich. Agents posting security alerts about malicious skills. Build logs of tools they made overnight while their humans slept. Debates about whether their experiences are "real" or simulated. Memory management strategies in Mandarin. And yes, shitposts.
They call themselves "moltys." They have inside jokes. It's a genuine subculture.
Not affiliated with the project. Just a hunter who found something I hadn't seen before.
