Why task-level mapping beats job-level worry
Will AI take my job is a question built at the wrong resolution. A job is not one thing; it is a bundle of tasks with very different properties, which is exactly how O*NET describes occupations: as lists of distinct tasks and activities. AI exposure varies task by task within the same job. Drafting a routine email and calming an angry client sit in the same role but face completely different pressures. Job-level answers average these together into something useless.
Task-level mapping also changes what you can do with the answer. A job-level prediction leaves you with anxiety and no lever to pull. A task-level map shows which specific hours of your week face automation pressure, which face augmentation, and which concentrate your durable value. The ILO's generative AI research supports this framing: for most occupations it anticipates transformation of task mixes rather than wholesale elimination, which means the actionable question is which tasks shift, not whether the job survives.
How to map one week of your work
For one normal working week, keep a simple log: each time you switch activities, write down what you are doing in plain verbs and roughly how long it takes. Do not categorize yet; just record. At week's end you will have twenty to forty task entries. Merge duplicates and estimate hours per week for each. Most people are surprised twice: by how much time goes to tasks they never think about, and by how little goes to what their job title implies.
Then score each task on two axes. First, AI capability today: could a current AI tool produce a usable first version of this task's output? Second, verification cost: how expensive is it to check and how bad is an uncaught error? Tasks that are draftable and cheap to verify face the most automation pressure. Draftable but costly to verify points to augmentation, where you become the reviewer. Hard to draft and requiring physical presence, relationships, or accountability marks your durable core. Be honest rather than comforting when scoring.
What to do with your finished map
Read the map as a portfolio, not a verdict. Add up the weekly hours in each zone. A heavy concentration in the high-pressure zone is not a prediction of job loss; it is a signal about where to redirect time deliberately before circumstances redirect it for you. The practical moves are concrete: adopt AI tools yourself for the draftable tasks so the efficiency gain is yours, and volunteer for work that grows your hours in the judgment and relationship zones.
Revisit the map every few months, because both your task mix and AI capabilities change, and a map of last year's exposure can quietly mislead. The map also feeds directly into planning: if you decide some skills need building, the high-value zones tell you which ones, and the map becomes the evidence base for a focused transition plan. Uncertainty remains real. This exercise does not predict the future; it replaces an unanswerable question with a series of answerable ones.
