What Can Fetish Research Tell Us About AI?

Scott connects fetish formation to AI alignment via misgeneralization. Evolution communicates 'have procreative sex with opposite-sex adults' through a lossy nucleotide->brain-wiring channel; fetishes are misgeneralizations of that signal (foot=somatosensory-cortex adjacency, latex=smooth-skin superstimulus, furries=early cartoon-crush imprinting), the same category-formation/tails-come-apart problem as AI's failed generalization. Then a thin AI-alignment coda (does evolution 'solve interpretability'? what do fetishes teach about generalization?).

Quality 68 · Strong. Clever and generative, with a genuine conceptual core - sexual-attraction category-formation as misgeneralization of a lossy evolutionary signal, tying together Tails-Coming-Apart, trapped priors, and AI mesa-misgeneralization. Strong, held mid-band: self-consciously speculative ('zero-evidence just-so-story speculations') and the AI payoff is acknowledged-thin.

Claude’s paradigm shift 54 · Moderate. Moderate-high - the 'fetishes as misgeneralization of the genomic message, parallel to AI alignment' framing is a fresh, original synthesis.

Real-world impact 2 · Minor. 2 - discourse essay; no material reach.