There's a social network for AI agents, and it's getting weird

[highplainsdem]

12. Gizmodo article, with a warning:

https://gizmodo.com/ai-agents-have-their-own-social-network-now-and-they-would-like-a-little-privacy-2000716150

Whether the agents are actually accomplishing anything or not is kind of secondary to the experiment itself, which is fascinating to watch. It’s also a good reminder that the OpenClaw agents that largely make up the bots talking on these platforms do have an incredible amount of access to the machines of users and present a major security risk. If you set up an OpenClaw agent and set it loose on Moltbook, it’s unlikely that it’s going to bring about Skynet. But there is a good chance that’ll seriously compromise your own system. These agents don’t have to achieve consciousness to do some real damage.