Will AI Replace Mathematicians?
Among the most exposed occupations measured, ranked 10 of 785. Employment is projected to stay flat to 2035, at around 2,000 jobs.
Where this job sits against every other
Each column is a slice of the 785 occupations Microsoft measured for how much of their work today's AI can do. Mathematicians sit in the blue column.
What the job is made of
The biggest single piece of this job is “Analyze scientific or applied data using mathematical principles”, at 40.0% of the work. The column on the right is a different measure entirely: out of every time someone asked AI to do that activity, how often it actually got it right. It counts attempts people made, so a high number does not mean the work has moved to AI. It means AI is capable of that task when asked.
8 activities make up this occupation in the government's classification. The six here are the ones that define it most.
AI is weakest at “Analyze scientific or applied data using mathematical principles”, which is 40.0% of this job. Asked to do it, it finishes correctly 58% of the time. The work it handles well sits above 90%.
The numbers behind it
Employment and pay are United States figures, the only country publishing this per occupation. The work activities and the AI measurements are not country specific.
There are about 0,000 openings a year here, counting people who retire or move on as well as new positions.
This page is about the occupation. Your job is more specific than that.
The calculator reads your actual role and the tasks you do. It takes about a minute.
Analyze my jobOther jobs people ask about
- AI exposure and work activity measurements: Tomlinson et al., Working with AI: Measuring the Applicability of Generative AI to Occupations, Microsoft Research. CC BY 4.0.
- Employment, projections and pay: US Bureau of Labor Statistics, Employment Projections, 2025 to 2035. Occupation code 15-2021.