Frequently asked questions
What to know before you read your result, and answers to the questions that come up most often.
What your badge says — and what it does not
The score shown is a share of tasks, not a probability of replacement. It answers one precise question: for what share of the tasks in your occupation does a generative AI cut the time to complete them by at least half, at equal quality? A score of 82 out of 100 means 82% of the occupation's tasks can be greatly accelerated — not that 82% of jobs disappear.
The distinction is not a wording detail. An accelerated task is a task you supervise rather than execute. The freed time moves toward what AI cannot do: judgement, relationships, accountability. That is exactly what the sheet breaks down, task by task, for the reference occupation.
Where the figures come from
Exposure comes from Eloundou et al., “GPTs are GPTs” (2023), the reference study on the subject. Wages and growth projections come from the U.S. Bureau of Labor Statistics (May 2024). Courses and pivot occupations are an indicative demonstration: they will be sourced and dated before the final launch.
How to use this badge
The badge has two uses. The first is personal: read the sheet, identify the tasks you can delegate, and pick the course that moves your time toward your added value. The second is professional: share your badge or export it to feed a conversation with your employer about how your role is evolving.
Keep the tool's honest limit in mind: the average score of an occupation predicts nothing about your role, your company or your market. It situates your occupation; it does not classify you.
My occupation has a high score: is it going to disappear?
No. The score measures the share of tasks that AI can accelerate by at least 50%, not the probability that your job disappears. A high score mainly means the occupation will change fast and a large part of the working time can be delegated. The sheet shows what stays human: judgement, relationships, accountability.
Where do the exposure values come from?
From Eloundou et al., “GPTs are GPTs” (2023), which estimates the exposure of more than a thousand occupations to large language models. Some occupations show the value published in the paper; the others carry a provisional estimate, clearly flagged on the sheet, until the full dataset is extracted.
Why are wages in US dollars?
Because the most stable and up-to-date source — the U.S. Bureau of Labor Statistics — publishes in dollars. These figures are used only to compute the relative wage differences between occupations (as percentages), which stay readable in any country. Local wages (Canada, France) will be added later.
How are the courses chosen?
They have a single goal: move your time from delegable tasks toward your added value. Each course carries a relevance score for the occupation shown, plus indicators of duration, difficulty, cost and certification. This is an indicative demonstration: links and costs will be sourced before launch.
How are the pivot occupations chosen?
They are chosen as natural extensions of the current occupation: same family of skills, comparable or higher wage, and an AI exposure that raises the professional's value rather than threatening it. The wage gap is expressed as a percentage, which makes it comparable across countries.
My occupation is not in the list: what now?
The current version covers about twenty reference occupations. The full referential — more than a thousand occupations from the Eloundou dataset — will be added gradually. Write to us to request an occupation: the most requested ones are added first.
Is the site free?
Yes. The tool is free and will stay free. The planned model is affiliate links on the recommended courses (the site may receive a commission if you sign up through a link, at no extra cost to you), plus an optional detailed PDF report. No data you enter ever leaves your browser.