factual

Blue banana test: verifying outcomes we can't understand

We can get rich empirical feedback on AI safety by setting verifiable challenges—e.g. giving an AI control of an air-gapped computer's keyboard and asking it to root the environment to make a blue banana appear—because we can see whether it succeeded even if we'd never have detected the particular exploit, letting us test whether our interpretability and adversarial methods catch an AI using its best efforts.

factualpending

Speaker

Carl Shulman

Evidence Quote

“This can give us a rich empirical feedback where we're able to identify things that are even an AI using its best efforts to get past our interpretability methods, using its best efforts to get past our adversarial examples.”

Source

Carl Shulman (Pt 2) — AI Takeover, bio & cyber attacks, detecting deception, & humanity's far future— Dwarkesh Patel Podcast
Created: 6/13/2026, 3:35:43 AM

Your notes

Loading notes...