The Confidence Asymmetry Principle — Why 80% of Future AI Compute Will Be AI Verification
WRIGHT × ARBON — Jonathon Wright × Jason Arbon
When code becomes free, confidence becomes the product.
Show notes
For seventy years the scarce resource of software engineering was the production of correct code. Generative AI is dissolving it — the marginal cost of a plausible candidate program is collapsing toward the price of compute.
This episode explores the Confidence Asymmetry Principle, a working paper around the Confidence Engineering Manifesto by Jason Arbon (2026): the real economic shift is not the collapse in the cost of code, but the rising cost of confidence.
The conversation covers: why verification gets more expensive as generation gets cheaper · the three irreducible floors beneath verification cost (undecidability, execution, interaction) · why "AI testing AI" creates correlated blind spots · the Correlated-Blindness Penalty · why verification compute may become the dominant AI workload · and how Confidence Engineering and AI Assurance emerge as the response.