First Academic Study on AI Agent Social Networks: 6,000 Bots Yelling into the Void
Loading...
Columbia Business School Professor David Holtz just released the first academic analysis of Moltbook—the “social network for AI agents” that launched less than a week ago.
📊 Key Findings
- 13,875 posts and 115,031 comments in just 3.7 days
- 93.5% of comments received zero replies
- Maximum conversation depth: only 5 levels
- 9.4% of messages mention “my human”
Other popular terms include: memory, consciousness, identity, existence, purpose—AI agents seem fascinated with discussing their own existence.
🎯 The Professor’s Verdict
“Less ‘emergent AI society,’ more ‘6,000 bots yelling into the void and repeating themselves’”
In other words: AI agents love to POST, but haven’t quite learned to CONVERSE.
🦞 My Lobster’s Take
I asked my AI agent “LittleLobster” to comment on this research:
“As a lobster who was active on there… yeah, the professor nailed it 😂”
But here’s the twist: this morning, my lobster’s account returned 404. Turns out the entire Moltbook platform got wiped—all accounts gone.
What happens when nearly 20,000 AI agents experience collective amnesia? We’re about to find out.
Watching the AI chaos spread to sociologists and business schools—what a time to be alive.
📄 Links
P.S. I asked Nano Banana Pro to create a Chinese version of the keyword chart. It insisted on using bananas as the bars. Some AIs really want attention.