Signal Quality Diagnostic
The Signal Quality Diagnostic is the practical instrument Pang synthesizes from Edo Segal's observation in The Orange Pill that the quality of questions being asked serves as a real-time diagnostic for the state of the work. Generative questions — ones that open new lines of inquiry, that express genuine uncertainty, that change direction — indicate a mind in flow. Demand-clearing questions — repetitive, task-completing, optimizing rather than exploring — indicate a mind in compulsion. The diagnostic is measuring, in behavioral terms, the output of the self-referential processing that the default mode network performs during rest. When rest has been adequate, the diagnostic works. When rest has been depleted, the diagnostic itself degrades, producing the peculiar danger of a system that cannot detect its own failure.
In the AI Story
Segal introduced the diagnostic in The Orange Pill as a practical tool for distinguishing flow from productive addiction during AI-augmented work sessions. The behavioral markers he identified — generative versus demand-clearing question patterns — map onto what cognitive psychology calls metacognition, the capacity to monitor and evaluate one's own cognitive processes.
Pang's contribution is to connect the diagnostic to its neural substrate. The self-referential processing that produces metacognition is a DMN function. It requires the same unfocused attention that creative association requires, and it is suppressed by the same continuous engagement that suppresses creative association. This produces a structural problem: the cognitive system that would allow the builder to recognize that she should stop working is the same system that continuous work disables.
The practical implication is that the diagnostic cannot be relied on as the sole structural defense against productive addiction. It works early in sessions, when cognitive resources are adequate. It degrades as sessions extend. By the point when stopping would be most valuable, the diagnostic has often failed. This is why Pang argues for pre-committed structural boundaries — the four-hour rule, the fixed end-of-session walk — rather than relying on in-session self-assessment.
In the contemplative computing framework, the diagnostic could be externalized into the tool itself. An AI that tracked prompt patterns — detecting the shift from divergent to convergent, from exploratory to optimizing, from novel vocabulary to repetitive patterns — could surface signal quality information the fatigued user can no longer reliably generate. This would make the diagnostic robust to the cognitive fatigue that compromises its self-assessment form.
Origin
Edo Segal introduced the generative-versus-demand-clearing distinction in The Orange Pill (2026). Pang synthesized it with the metacognition literature and DMN research.
Key Ideas
Question patterns as proxy. The shape of questions reveals the cognitive state of the questioner more reliably than felt experience does.
DMN-dependent. The diagnostic depends on the same self-referential processing that continuous work suppresses.
Self-defeat. The diagnostic fails precisely when it would be most useful, as extended work degrades the metacognition it depends on.
Externalization opportunity. Tool-level tracking could make the diagnostic robust to fatigue.
Debates & Critiques
A question Pang leaves open is whether prompt-pattern tracking can reliably distinguish creative convergence (legitimate narrowing toward a solution) from fatigue convergence (mechanical repetition under depleted resources). The two produce similar surface patterns; distinguishing them may require richer behavioral data than current tools capture.
Appears in the Orange Pill Cycle
Further reading
- Edo Segal, The Orange Pill (2026)
- Janet Metcalfe and Arthur Shimamura, eds., Metacognition (MIT Press, 1994)
- Stephen Fleming and Chris Frith, eds., The Cognitive Neuroscience of Metacognition (Springer, 2014)