Dunning Kruger Misalignment

Some people with little skill think they are very good because they cannot see their own errors. This leads them to act confidently even when their choices are poor.

A cognitive bias where low-competence agents overestimate their ability due to limited metacognitive insight and feedback, producing systematic miscalibration. This miscalibration can distort decision-making and performance evaluations within information selection ecosystems.

A new employee who has only skimmed a company manual confidently rewrites the filing system, certain they've improved it, while a seasoned colleague who knows the system's nuances stays quiet — and the overhaul quietly breaks dozens of dependent processes nobody notices until much later.

In a digital archival ingestion pipeline, a junior archivist with limited metadata schema knowledge applies heuristic shortcuts to a large record batch, assigning high retention priority to visually prominent but substantively redundant documents while discarding ephemeral correspondence containing unique provenance metadata. Because the archivist's metacognitive monitoring is weak, the miscalibrated appraisal decisions are never flagged; high confidence scores in the cataloging interface suppress escalation triggers, producing systematic skew in what gets ingested. Downstream retrieval modeling inherits these skewed acquisition thresholds, and the archival gaps created by the discarded records are never surfaced in finding aid completeness audits — a textbook case of overconfidence compounding selection bias within curation workflows.

When someone cannot judge their skill, they assume their answer is right and stop looking. That overconfidence causes poor choices to be repeated and shared.

A weighting_asymmetry arises because limited metacognitive monitoring reduces negative feedback weighting, while partial competence nodes maintain higher corrective signal gain; the archival indexer and curator modules thus propagate overconfident but low-quality selections. Structural constraint in evaluative pathways causes asymmetric error amplification and undercorrection across selection layers.

Give clear, simple feedback about specific errors and facts. Make people compare their answers to reliable examples so they see mistakes.

Implement objective performance benchmarks and calibrated feedback loops that expose discrepancies between confidence and accuracy, integrating peer review and ground-truth audits. Adjust selection weighting to downrank high-confidence low-accuracy contributions through automated validation gating.

Overconfident low-quality curation; False consensus formation; Feedback signal neglect

An adversarial actor can deliberately suppress corrective feedback channels — for example, by designing curation or peer-review pipelines that reward confident submission volume over accuracy — so that low-competence agents dominate selection outputs while remaining unaware of their miscalibration. This can be weaponized in information ecosystems to flood archival or curatorial layers with overconfident low-quality content, crowding out high-skill contributions whose authors self-censor due to perceived uncertainty. Adversaries may also exploit the bias by positioning overconfident proxies as authoritative voices in high-stakes decision contexts, knowing those proxies will resist corrective feedback and sustain the distortion.

Implement calibrated confidence elicitation protocols (e.g., asking agents to provide probability estimates alongside selections) and expose discrepancies between stated confidence and objective accuracy benchmarks through automated validation gating. Mandate structured peer review with ground-truth audits and require contributors to engage with explicit error-rate data before their selections are propagated. Longitudinal tracking of individual selection accuracy versus confidence scores can surface persistent miscalibration patterns and trigger targeted remediation.