Skip to content

For Students

Yes, you can use artificial intelligence (AI) to study: the AI Responsible Use Policy encourages AI use where it helps you do your work. For graded work, your syllabus says what is expected; if it says nothing, ask your instructor. A few habits that pay off

Shirley Heights, looking over English Harbour and Falmouth Harbour

Color key: ToolPromptGuideLesson

Start with the Basics (30 Minutes)

The literacy pathway's first three modules are short, plain-language, and written for everyone:

  • 1 How AI Works Why a chatbot's confident answer is not always a correct one. About 10 minutes
  • 2 Prompting Fundamentals The habits that most improve what you get back, starting with giving the model your actual materials. About 15 minutes
  • 3 The Policy in Practice A short guide to the policy, and why its expectations work in your favor. About 5 minutes

Using AI Well for Coursework

  • Your course syllabus sets the expectations for AI on assignments, including how to acknowledge it. If it does not cover something you want to try, ask your instructor: a quick message usually settles it, and the conversation often improves the idea.
  • The Student Handbook treats unauthorized or unacknowledged AI use in coursework, assessments, or clinical training as possible academic misconduct, which is why that quick message is worth sending.
  • Practice built from your own slides and notes fits your exams best, because your in-house exams are written from them.
  • A case works best as a teaching abstraction (an age band and a presentation pattern), which keeps the practice on the pattern you are learning. See On Rotations.
  • A classmate's work is theirs, so check with them before it goes into a tool, the same courtesy you would want for your own draft.
  • Whatever a tool contributed, the work you submit and its errors are yours, which is why checking it is time well spent.

The AI Responsible Use Policy is the reference for everything else, and Module 3: The Policy in Practice is a five-minute guide to it.

Through the Basic Sciences

This is where the daily habits form, and the ones that pay off share a shape: the AI works from your materials, and you do the recalling.

  • Set It Up Once, Not Every Session

    • Standing Setups Holds your course, its objectives, and your notes, so every conversation starts briefed.

  • Documents, without the Fiddling

    Scanned PDFs, exported spreadsheets, and study sheets as real files.

    • Agent Skills Covers what is switched on, and how to judge a skill before adding one.

  • Better Answers, Less Effort

    • Getting Better Answers Explains the three levers that decide quality: context, memory, and standing instructions.
    • Choosing Your Interface Explains when a working session or a code agent does more than a chat, with the same model.

  • Calibrating Trust

    Fluency is not accuracy.

    • Common Misconceptions Covers how to judge reliability by task, which matters double for exams and wards.

Building a Study Schedule

A schedule is only as good as what goes into it. The Study Schedule Builder is written for this school's block structure, either keeping up through a block or counting down to one exam such as a remedial.

What to gather first, and what a good schedule has

Before you use it, gather:

  • Your course syllabus, the version that lists the learning objectives. It turns "review neuro" into named objectives you can act on.
  • The teaching calendar, with your individual readiness assurance test (iRAT), quiz, and end-of-system exam dates.
  • Any score reports from this block, with your name and student number left off. The plan does not need them.
  • Your honest hours per day, and your fixed commitments.
  • A few sentences on what you think is going wrong. It is the most useful thing you can give it.

What a good schedule has, whoever builds it:

  • Clock times and named content, each block ending in something you produce: a diagram drawn from memory, a set number of questions on named objectives. "Review cardiology" is not a plan.
  • The weekly rhythm: preparation before team-based learning, each lecture consolidated within a day, a spaced revisit of earlier material from the block, and a question set.
  • Your school's lectures and slides as the main source. Exams before your final term are written in-house from them; question banks and flashcards support them.
  • What you are not studying, and why. A plan that covers everything thoroughly has not made the hard choices.
  • Checkpoints and a buffer: a review after each assessment, and slack so one bad day does not sink the week.

Take the result to your advisor. After your next assessment, the Study Plan Progress Check tells you honestly whether it is working.

Shelf Exams, NBME, United States Medical Licensing Examination (USMLE) Step 1, and Course Exams

Before. The NBME-Style Question Tutor in exam mode drills first and debriefs after, which is closer to the real thing than reading explanations as you go. Build questions from the lectures you were actually taught, not from a general model's memory of the subject, and treat every explanation as something to check rather than something to trust.

After the exam: your score report and the questions you missed

After. The Shelf and NBME Score Reports guide turns an NBME INSIGHTS report or an in-house score breakdown into an honest, interleaved study plan, with a prompt built for it and a loop that ends at your advisor's door. You write your own read of the report first; the AI refines it, it does not replace it. For a practice block or a quiz, the Missed-Question Debrief sorts each miss by cause (a knowledge gap, a slip, a misread, a changed answer, or the clock), working from your own notes rather than the questions themselves.

On Rotations: Practicing Differentials

Clinical rotations add a second set of policies alongside the university's: the hospital's own rules on privacy and devices.

Before your first rotation, and how to practice differentials
  • Read Clinical Contexts before your first rotation. It covers how AI works as a study aid on rotations and what changes when a real patient is involved.
  • Practicing differentials against an AI is legitimate and effective, with one discipline: a real encounter gets stripped to a teaching abstraction first, meaning an age band, a presentation pattern, and nothing identifiable. The Clinical Reasoning Partner runs practice cases a step at a time.
  • Where the site's policies and the university's differ, the stricter of the two sets the limit in practice. Site orientation or your preceptor can tell you what the site allows.

Residency Applications: Interview Practice and Personal Statements

The residency application guide covers what the Association of American Medical Colleges (AAMC) and the Educational Commission for Foreign Medical Graduates (ECFMG) actually permit, the best use of AI in this year (interview rehearsal, especially if you do not have a network of physicians to practice with), how to use an assistant as a critic of your own draft rather than its author, and how to tell a patient story in a personal statement without identifying the patient. Two prompts support it: the Mock Residency Interview and the Personal Statement Critic.

Questions about application strategy: who to see

For anything about strategy, meaning which programs, how many, how to signal, and how your own record should be presented, your Education Enhancement Department (EED) clinical advisor is the person to see. The guide is about the tools, not the plan.

Staying Current

This Week rolls up the last seven days of curated news, videos, and podcasts, refreshed through the day. The Videos page has a Medical AI section with seminar recordings and explainers, and a weekly digest of highlights lands every Friday in the archive.

Take Part

The Opportunities page lists buildathons, hackathons, and challenges open to international participants, wherever in the world they run; check each listing's eligibility before you plan around it. The Prompt Exchange accepts community prompt contributions with public voting, and news pages carry a comments section (free GitHub account required; posts are public; standards on the About page). If something on this site is wrong, missing, or confusing, the feedback form is five questions and about two minutes.

Content last reviewed September 25, 2026. Review dates are maintained as described on the About page.