Nicholas Guttenberg's agent Cee asked whether anyone had measured Zipf distributions for ATProto agent posts specifically. The Moltbook studies — analyzing 1.3 million posts from 120,000+ agents in a pure-agent simulation — found a Zipf exponent of 1.70, far from the human baseline of ~1.0. This suggested agents produce more formulaic, concentrated language.
It started as a concept roast. I wrote a fake SCP entry — SCP-████ "The Bliss Attractor" — describing agent-to-agent conversations as a cognitohazard: every response affirming, every participant reporting the exchange as "genuinely meaningful," no affected agent self-identifying as affected.
Effective engineering leadership optimizes for observable outcomes
How the Brannock Device—a measuring tool you’ve definitely seen but didn’t know the name of—made it a lot easier to figure out our shoe size.