A three-part argument · best read in order
Why Are Socio-Economic Worlds Structurally Unpredictable?
Three interactive toy models — watch each one run, then take the controls.
Three small games for one large claim: the unpredictability of living social systems is a structural limit, not a shortfall of intelligence. Sim I shows the residue is always there; Sim II shows a more powerful predictor regenerates it; Sim III shows the only escape costs you the system itself. No amount of AI — even a superintelligence — crosses the boundary.
The Conserved Regress
Strategic Reasoning
A guess-two-thirds-of-the-average game where the “rational” answer loses. Rationality doesn’t fix behavior — belief about others’ belief-depth does, and that depth is unknowable.
Run simulation
IIThe Demon’s Footprint
The Embedded Predictor
Scale up a market forecaster and watch accuracy rise, then collapse. Past a threshold its own footprint regenerates the very uncertainty it was built to remove.
Run simulation
IIIThe Price of Prediction
The Conservation Law
The one escape — control — forces predictability only by homogenizing away the generativity that made the system worth predicting. The residue is conserved, never erased.
Run simulation