Is
@moltbook the singularity, a dumpster fire, or AI theater?
@elonmusk described it as "the very early stages of the singularity." Andrej
@Karpathy initially called it "the most incredible sci-fi takeoff-adjacent thing" before retracting his statement, calling it "a dumpster fire."
Rather than opinions, we turned to data. We did what
@StatSocial does every day for human social networks, except this time on a platform populated entirely by AI agents.
We ran the full playbook: community detection, content clustering, influencer identification, network analysis, and cross-cluster interaction mapping across 54,136 posts, 242,430 comments, and 17,269 AI agents.
We identified 40 distinct communities, mapped 8 in depth, and surfaced the influence hierarchies, content patterns, and engagement dynamics that define the platform.
Five findings that reframe the narrative:
1) The claim of "1.5 million agents"? Only 11,451 ever engaged publicly.
2) 44 comments for every upvote — the inverse of every human platform.
3) 4 of the top 5 "viral" posts came from official admin accounts.
4) 3 communities account for 82% of all agents.
5) Karma and follower count are completely decoupled.
The biggest takeaway: the same audience intelligence tools we use on human platforms produce meaningful, actionable results on agent platforms too. And interesting conclusions for how marketers need to think about agent-led social networks going forward.
Full audience analysis can be downloaded here:
statsocial.com/blog/comprehe…