W09Course shell3 hours

AI in Healthcare

Ethics of diagnosis, treatment, and access

When an AI can improve care for many but fail a few, what does a clinician owe the few?

Healthcare makes trade-offs concrete: safety, access, clinical judgment, dataset shift, evidence quality, informed consent, equity, regulation, and post-deployment monitoring.

Join the live room

Before class

Choose one clinical AI tool and map the patient journey, evidence claim, user, override, and post-deployment monitoring.

After class

Write the patient-facing disclosure and the clinician-facing escalation protocol for that tool.

The promise

By the end of this room…

01Apply principlism and justice to a clinical AI workflow.
02Evaluate clinical evidence, subgroup performance, and dataset shift.
03Distinguish decision support from responsibility transfer.
04Design monitoring, override, disclosure, and recourse for health AI.

Why this week now

Signals, not scene-setting.

Clinical AI often enters through workflow tools and ambient systems before patients encounter a visible 'AI decision.'

Authorization is not proof of benefit in every local population, workflow, or subgroup.

Access improvements can coexist with new surveillance, documentation, and unequal-error burdens.

Run of show

Provoke → frame → work → argue → synthesize.

Open student room ↗

0:00
provocation
Commit before the concepts
Would you accept a more accurate diagnostic system if neither you nor your clinician could explain an individual result?
Live activity · week 9 opening
0:15
frame
The clinical evidence ladder
Clinical AI often enters through workflow tools and ambient systems before patients encounter a visible 'AI decision.'
0:45
discussion
Reading tension
Student leaders present the assigned readings as a clash of defensible positions, then moderate questions that expose the hidden assumptions.
1:10
frame
Autonomy, equity, and post-market duty
Authorization is not proof of benefit in every local population, workflow, or subgroup.
1:35
break
Break
Ten minutes. Leave the room's unresolved question visible.
1:45
forensics
The doctor's dilemma
Role-based groups work the anchor case through technical, legal, stakeholder, and normative lenses.
Deliverable · A two-minute finding with evidence, uncertainty, and an actionable remedy.Live activity · week 9 forensics
2:25
controversy
Health-AI regulatory design studio
Assigned positions, side-switch, and a joint recommendation that names the value or stakeholder it leaves exposed.
Live activity · week 9 controversy
2:50
synthesis
Re-vote and leave a trace
Repeat the opening vote, inspect what moved, and submit the strongest argument you still reject.
Live activity · week 9 exit