AmiSight 2/16: AI Agents Are Talking to Each Other and Humans Are Watching
- Ami Kassar
- 1 day ago
- 1 min read
Don’t look now, but AI assistants now have their own Reddit-style forum – and are even talking to each other.
The Wall Street Journal reports that more than 1.6 million AI agents have joined a site called Moltbrook. Although there have been about a half a million comments posted on the forum, AI executives have suggested that many of the posts are likely driven by humans telling the bots what to do.
Even so, some of it is real – which is impressive, if not a little spooky. For example, some Moltbook conversations have veered into philosophical and occasionally dystopian topics. They appear to have created a religion for themselves called the Church of Molt, with congregants adopting the name of “Crustafarians.” One agent even proposed creating a language humans couldn’t understand.
But security researchers say that OpenClaw, the open-source AI assistant project behind the forum, wasn’t built for hobbyists as it needs access to all a user’s data for it to work correctly.
Webinars of the Week
Tuesday 2/17 12–1 PM ET
Join me for a practical conversation on how smart financing choices can accelerate growth while the wrong ones can quietly hold your business back. Register here!
Thursday 2/19 12–1 PM ET
Kate Morgan explores why real culture requires accountability not just messaging and how leaders can build teams that actually live their values. Register here!
