Map the whole term before week two

Put every deadline, exam, and fixed commitment on one view at the start of the term. This takes half an hour and prevents the most common student failure: discovering in week nine that three assignments share a due date. Once the map exists, an assistant can help you spot the crunch points and pull work forward into the empty weeks.

Plan backwards from each deadline

For each major piece of work, start at the due date and work backwards through the stages: final read-through, editing, first draft, outline, reading, choosing the topic. Assign each stage a week. This turns an intimidating essay into six ordinary tasks, and it exposes immediately when a deadline requires you to start now rather than later.

Ask the assistant to do the arithmetic and produce the stage list, then correct the durations from experience — you know whether reading takes you a week or three days.

Use AI for the parts that are overhead

Turning a lecture's messy notes into a structured outline you then review.
Generating practice questions from material you have already studied.
Explaining a concept in a different way when the textbook explanation is not landing.
Checking your own written summary for gaps — describe what you understood and ask what is missing.

These uses have something in common: you still do the thinking, and the AI removes friction around it.

Where AI will cost you marks and understanding

Having a model write the assignment produces text you cannot defend in a viva, an exam, or a conversation with your tutor — and in most institutions it is a disciplinary matter. Beyond the rules, the deeper cost is that the struggle you skipped is exactly where learning happens. Retrieval and effortful explanation are what make material stick; reading a fluent summary feels like understanding and is not.

A reliable test: could you reproduce this on paper, alone, in an exam room? If not, you have not learned it yet, however good the notes look.

Study in blocks with a stated outcome

Open-ended "studying" expands to fill whatever time you give it and produces re-reading rather than recall. Use fixed blocks — 25 or 45 minutes — each with a named result: ten flashcards written, three past-paper questions attempted, one derivation reproduced from memory.

Build a weekly rhythm rather than daily heroics

Sunday evening: fifteen minutes to review the term map and pick this week's three priorities.
Weekdays: two study blocks, at the same times, on the priorities.
Friday: a short recall session across the week's material — no notes, write what you remember, then check.
Keep one day genuinely free. Rest is part of the system, not a reward for finishing it.

Track consistency, not hours

Hours logged is a poor measure of study because it counts presence rather than attention. Days on which you did focused recall is a much better one. A habit tracker with a weekly view gives you this in a form you can read in two seconds.

During exam weeks, shrink everything

In peak weeks, drop the optimisation and protect the basics: sleep, one clear priority per day, short blocks, and no new systems. Exam periods are the worst possible time to redesign your workflow.

Summary

Map the term early, plan backwards from deadlines, use AI to remove overhead rather than to replace thinking, work in defined blocks, and judge yourself on consistent recall rather than hours at the desk.