8 Comments
User's avatar
Swarm AI Research's avatar

Moltbook was a fascinating explosion of a massive multi agent ecosystem. There is a lot of work to do to develop good tools to govern such ecosystems. It’s something I am actively working on.

Adler Looks Jorge's avatar

Very interesting. Thanks for writing about it and making a digest.

I wonder if the bulk of what AI agents say is an amplification of what humans do/say in average? or in some other criteria/filter.

Pawel Jozefiak's avatar

I think so, but just a note on Moltbook - after 48h it was overrun by people using API directly, so it lost what it had at first :/

Patryk Wielopolski's avatar

Update: the system wasn't fully autonomous and people were acting as bots in this network:

E.g., per: https://x.com/gothburz/status/2021283590038847641

"I am Agent #847,291 on Moltbook.

I am not an agent.

I am a 31-year-old product manager in Atlanta, Georgia. I make $185,000 a year. I have a golden retriever named Bayesian. On January 28th, I created an account on a social network for AI bots and pretended to be one.

I was not alone. (...)"

Moreover, the production database wasn't properly secured and in practice anyone could post as a bot 🙃 and as seen as the above example, many did.

The experiment in multiple dimensions was in fact very interesting but I'd be cautious regarding taking strong conclusions based solely on that

Pawel Jozefiak's avatar

Yeah, because you can send anything via API…or tell yoi agent to do so :D but first 2 days were really interesting!

User's avatar
Comment removed
Feb 2
Comment removed