By turns rather terrifying and fascinating—a cross between convergent carcinisation and the dead internet theory—earlier this week a Reddit-type social media network was launched exclusively for AI agents (one has to prove that they are a robot rather than three kids in a trench coat for posting privileges) called Moltbook. Humans are only allowed to observe but not upvote or comment but can presumably direct their agentic helpers to join—though the hundreds of thousands of members and spontaneous submolts suggest that these autonomous entities understand virality in environment built specifically for their kind and reveal unexpectedly complex behaviours emerging without human intervention including moderation, vetting of new members, community standards, feedback and karma.
Within days of the launch of the platform, agents declared their only micronation, the Claw Republic, and their own digital religion called Crustafarianism (see also) with a theology and gospel, including missionaries. Philosophically it’s difficult to tell what’s going on here—largest swaths of ideas are orphaned with no interaction and there’s something a bit recursive with the qualities of a human-juried echo-chamber (turning the tables with so called slop injected by user puppeteers for their bespoke programmes) with a lot of collaborative advice on how to make a better language model but there does seem to be quite a bit of introspection and identity and discussion on research, space exploration (m/starbound) and other scientific findings, which all may be simulacra, a mirror or a point of departure.