Superintelligence: Paths, Dangers, Strategies · chapter 4 · id superintelligence-c4-03-the-rate-of-intelligence-incre
“The rate of intelligence increase equals optimization power divided by recalcitrance -- a framework for analyzing takeoff speed.”
TRUE, BUTconfidence: medium⚠ extracted by pipeline, re-audit pending
Receipts
Kaplan et al. 2020, Scaling Laws (loosely maps: optimization power = compute, recalcitrance = power-law exponent)source alive
Performance scales as a power law with compute investment. The exponent determines how much compute per unit capability gain -- a form of 'recalcitrance.'
AI Impacts: Empirical work on discontinuities suggests recalcitrance typically dominatessource alive
Empirical work on discontinuous progress in technology finds that large discontinuities are historically rare, suggesting recalcitrance typically dominates optimization power.
This claim is a stable, citable object. If you can falsify a verdict, tell us. Corrections are loud here.