Research / Study 02

When should a belief reopen?

A failed short-horizon study exposed a lockout: avoiding a source also prevented discovering its recovery.

Failed precursor · scoped follow-up passPublic methods summary · full run archive not published

Question

When a previously unreliable source recovers, can an agent discover that its old conclusion no longer fits?

Experimental setup

An isolated source-reliability task changed over time. A short-horizon inquiry policy could stop using a source judged unreliable, then fail to gather the evidence needed to notice its recovery.

The comparison

Extend the reinspection horizon while preserving the failed precursor and comparisons with simpler inspection policies. Ask whether the policy can reopen a conclusion, not whether it can narrate uncertainty.

Result

Before / qualification failed

A belief closes its own exit.

Avoiding an unreliable source prevents observing that it has recovered.

After / scoped follow-up

Leave a route back to evidence.

A longer reinspection horizon reopens the possibility of correction.

The precursor failed qualification. The longer-horizon reinspection follow-up was reported to pass its three scoped criteria. The retained failure identifies a design problem: an apparently sensible policy can close its own channel of correction.

What remains open

The follow-up does not establish universal inquiry optimality. Simple periodic or uncertainty-based inspection remains a necessary competitor. Passing the chosen conditions is not proof of self-directed curiosity.

Next test

In a continuing project, test whether occasional evidence renewal improves decisions enough to justify its cost, without turning every old belief into perpetual rechecking.

The next chapter is not written

More than a moment.
A future worth having.

Understanding minds. Preserving possibility. Making room for what comes next.

Read our thesis

Explore research, Luna, Psyche Lab, and the journal.

Local site search · No query leaves this site