AI slop is building a maze that even the smartest engineers can't escape
The Decoherence Trap
The essay introduces "decoherence": the buildup of locally valid but mutually incompatible models inside a system, until the cost of change is dominated by translating between them and dodging side effects. It grows faster the more decoherent a system already is, producing runaway exponential growth. AI is now generating it at an unsustainable rate. The upside: this rules out a runaway intelligence explosion. The downside: we are heading toward Decoherence Collapse.
When pressure is applied, the expert doesn't do the full analysis faster; they just stop earlier and do something they're pretty sure will work.