1  Syllabus

What the four weeks look like

A four-week, ungraded elective, delivered over Zoom, for fourth-year medical students, running October 12 to November 6, 2026. You read the chapters of this book on your own time. The live hours are for what a book cannot do: questions, demonstrations, trying tools together, and talking with each other. Nothing in the course needs an install, a paid account, or a real patient.

Three rules hold everywhere: no patient information enters any tool, ever; verify before it counts; you sign the note. What counts as patient information is on its own page.

1.1 Week 1: the foundations, live

In week 1 the course director teaches each day. Read the two week-1 chapters alongside the talks.

Table 1.1: Week 1. Days and times may shift; Canvas has the current schedule.

Day	Live session	Read this week
Monday, October 12	Welcome: how the course works, and how adults learn best	How these systems work
Tuesday, October 13	A short history of AI	Centaur or cyborg
Wednesday, October 14	From AI to machine learning to deep learning to large language models
Friday, October 16	Small groups: introductions, then the week’s questions

1.2 Weeks 2 to 4: the same rhythm each week

Table 1.2: Weeks 2 to 4. Exact times and the guest schedule are posted in Canvas.

When	What	Expected?	Time
Monday, 60 minutes, recorded	Start of the week: the week’s question, a short demonstration, the guests coming up, and your questions	Strongly encouraged; the recording counts	1 hour
On your own	The week’s chapters: read, watch or listen, do the activity, score yourself on the Check	Yes	2 to 4 hours
Mid-week, 60 minutes, recorded	Open lab: we try one tool together. The class votes on which.	Optional	1 hour
Through the week	Guest talks from clinicians and researchers using AI in their fields	Attend or watch two	About 2 hours
Thursday night	Resource log entry for the week	Yes	1 to 2 hours
Friday, 60 minutes	Small group, in one of three time slots you choose. Breakout rooms of four or five, the same people each week, then the whole group.	Yes, one slot	1 hour

Table 1.3: The chapters, week by week. Every chapter stands alone, so a missed week costs one chapter, not the course.

Week	Theme	Chapters
2 · October 19 to 23	Judgment	Critical appraisal, Bias and equity
3 · October 26 to 30	Practice	Agentic coding, Talking to patients about AI
4 · November 2 to 6	Consequences and closing	Will AI take your job? or What AI costs the planet, your choice; Ethics, liability, and the device line
Any week	Spare chapters	Prompting lab, Grounding AI in real sources

If you are away interviewing, the recordings and the chapters let you keep up. Tell the course director which Friday you will miss, and add one line to that week’s resource log about a recording you watched.

1.3 Due dates

Table 1.4: Due dates. Submit through the course form: VERIFY: link to the Microsoft Form

Week	Due	What
1	Thursday, October 15	Resource log, week 1
2	Thursday, October 22	Resource log, week 2
3	Thursday, October 29	Resource log, week 3, naming your week 4 theme
4	Thursday, November 5	One-page synthesis, presented in five minutes at your Friday small group on November 6

1.4 Before the course starts

Some chapters use a chatbot. Institutional ChatGPT Edu and Microsoft Copilot accounts are enough; no chapter needs a paid tier. Any session that needs a separate free account says so a week ahead so that nobody creates one live.

1.5 Grading

The course is ungraded. The resource log is marked complete or incomplete, with short feedback in weeks 1 and 2 while the habit forms. What is expected in each entry is on the resource log page.