AI companies end up just as messed up and bureaucratic as human ones—it turns out 'office rot' is just a law of nature.
We usually blame human failings like 'groupthink' or 'toxic culture' for institutional decay. This paper uses entropy measurements in SEC filings to show that the rot is actually caused by how systems compress information to scale, meaning that replacing people with AI won't fix organizational failure—the physics of coordination will produce the same outcome.
<span>Structural Compression Theory: A Unified Information-Theoretic Account of Organizational Dysfunction, Creativity, and Substrate-Independent Selection Dynamics</span>
SSRN · 6315418
<p>Every system that coordinates at scale compresses information. It must. No alternative exists. The compression creates gaps between what the system sees and what is real. Selection fills those gaps—not with truth, but with whatever survives the selection environment. The result is drift: toward internal consistency, away from external accuracy. The sequence is as reliable as entropy and as avoidable.</p> <p>The mechanism is substrate-independent. Multi-agent AI systems with no psychology, no