PAN Lab example

The same AI with a human checking: the supervised office

Same AI — but a human checks every output before it reaches the record. In the sociotechnical simulation — a modeled office, not a real one — that holds errors to about 20%, a quarter of the hands-off rate. The failure here is slow: the review keeps happening while the judgment behind it quietly thins.

Comparative teaching networkCaseworker documentation & copilots

Open this example in PAN Lab v0.1 to apply pressures and levers and watch what the system does.

What this models

This example runs on the Human-supervised office network: 3 components and 9 pathways between them. Every context in the Lab is a stylized model, never a reconstruction of any actual deployment, and each assumption behind it carries a provenance label.

Evidence base: 3 assumed · 1 published baseline. In the Lab, the shaded evidence band behind each headline readout draws its width from the least-established class below.

baseline
How often errors stick in this office (≈20%) comes from the sociotechnical simulation, not from measuring a real office.
assumed
Peer pathways are authored on both signs: answers spread collegially and reviewers compare notes; the shared assistant's errors are treated as correlated across the team (a monoculture assumption, not a measurement).
assumed
The model's raw error output is held identical across all three office cultures; only oversight differs.
assumed
Review quality is assumed steady over time; the deskilling drift explored elsewhere is switched off at baseline.

Sources and evidence

What this example rests on, claim by claim. Every entry resolves to the same ledger the Evidence Registry publishes.

In the sociotechnical simulation, the same AI in three modeled office cultures - stylized, not real workplaces - let errors stick at very different rates: roughly 75% under low-oversight autonomy, 20% under human supervision, and 16% under high-governance professional controls.
scenarioillustrative PAN-run result
No published source is attached to this claim yet.

Where this connects

Institutional pressures in this domain

Workload surge — Demand outruns staffing; per-case attention shrinks and review becomes triage.
Deadline pressure — Statutory or managerial timeliness rules reward fast approval of machine output over slow disagreement.
Reviewer bottleneck — One fixed-capacity checking stage sits between AI output and consequence; everything queues behind it.
Staff turnover — Experienced skepticism leaves; new staff calibrate their trust on the tool itself.
Vendor opacity — The deploying institution cannot inspect the model, data, or update pipeline it is accountable for.
Compliance over substance — Paper controls (sign-offs, checklists) satisfy audits while the behavior they describe erodes.

All of them in context on the Caseworker documentation & copilots domain page.

Levers available here and the patterns behind them

Upgrade model — Improve the model
Review the riskiest first — Risk-tiered oversight
Keep prompts neutral — Framing and mirroring reduction
Escalate checks — State-feedback vigilance
Keep skills sharp — Deskilling-arrest mandate
Understand the system — Understand the system
Mark AI-written records — Provenance labeling
Review on schedule — Oversight cadence & retrospectives
Store less data — Data minimization
Assign a challenger — Structured dissent