Preparing content…

Mathematician

SOC 15-2021 · O*NET 15-2021.00

Fundamental mathematics researchModeling and numerical computationApplications and methods
58%AI Impact ScoreFermatMind / 费马测试

Mathematicians frame questions that can be studied by proof or computation. They test conjectures with definitions, examples, counterexamples and models, and document the conditions under which results hold.

Model impact index for three equally weighted representative core tasks; not a measured share of working time or a job-loss probability. floor((50+75+50)/3+0.5)=58; fermatmind-ai-impact-v1; 2026-10-18.

+0.6%U.S. projected employment change (2025–2035)U.S. Bureau of Labor Statistics
Source: U.S. Bureau of Labor Statistics|United States|2025–2035|Projected employment growth; Editorial conversion: BLS Table 1.2: U.S. occupation 15-2021 had about 2,200 jobs in 2025 and is projected to have about 2,200 in 2035 (+0.6%), with about 100 openings per year. Counts are rounded independently; projected openings include replacements and do not measure current hiring.
About 2,200 jobsU.S. jobs (2025)U.S. Bureau of Labor Statistics
Source: U.S. Bureau of Labor Statistics|United States|2025|Employment count in 2024; Editorial conversion: BLS Table 1.2: U.S. occupation 15-2021 had about 2,200 jobs in 2025 and is projected to have about 2,200 in 2035 (+0.6%), with about 100 openings per year. Counts are rounded independently; projected openings include replacements and do not measure current hiring.
About 100 per yearU.S. annual openings (2025–2035)U.S. Bureau of Labor Statistics
Source: U.S. Bureau of Labor Statistics|United States|2025–2035|Projected job openings (period total); Editorial conversion: BLS Table 1.2: U.S. occupation 15-2021 had about 2,200 jobs in 2025 and is projected to have about 2,200 in 2035 (+0.6%), with about 100 openings per year. Counts are rounded independently; projected openings include replacements and do not measure current hiring.
No data availableMainland China pay reference

Quick assessment

Who might enjoy working as a mathematician?

Mathematicians focus on advancing mathematics and applying it to practical problems: researching traditional fields such as algebra, geometry, probability, and logic to expand mathematical knowledge; developing new principles and relationships among existing principles; formulating hypotheses and deriving their consequences; conducting numerical analysis and computation; building mathematical models for analysis or computational simulation; applying mathematical theory and techniques to practical problems in business, engineering, science, and other fields; developing computational solution methods; communicating findings through reports, papers, and conferences; and instructing others in mathematical methods.

The following assessments concern how mathematicians research and model. They do not predict hiring, qualifications, authorization, or research results.

Who is more likely to fit Mathematician?

Sample output and reflection: Check whether you first defined the problem and assumptions, explained methods and applicability conditions, provided verifiable conclusions and limitations, and clearly distinguished proof from numerical approximation. Then ask whether you can sustain derivation and computation, persist through long periods without results while honestly acknowledging uncertainty, and take responsibility for conclusions and citation standards.

When should you be cautious about choosing Mathematician?

This is an entirely fictional research scenario check, not a personality diagnosis, hiring conclusion, or qualification assessment.

How can a small experiment test your fit for Mathematician?

Entirely fictional task package: Provide a fictional problem context and data, including 1 problem description with constraints, 1 set of available observations or simulation data, and 1 modeling or proof objective. Within 60 minutes, write modeling assumptions and a mathematical formulation, a solution or proof approach, verification methods and limitations, and explain how the results can be tested. Do not use real confidential data or others' unpublished findings.

Career profile

Mathematicians frame questions that can be studied by proof or computation. They test conjectures with definitions, examples, counterexamples and models, and document the conditions under which results hold.

Quick career answers

What can a small mathematics research sample show?

State a question, define the objects and test a conjecture with examples or a counterexample. Separate what you have proved from what still needs proof.

Objects:

  1. 01Mathematical problems and models, data and computation

Inputs:

  1. 01Assumptions, method applicability, and standards
DeliverablesVerifiable proofs and findings

How does mathematical research differ from statistics?

Mathematical research often investigates structure or conditions for a result. Statistics often uses data and models to make inferences under uncertainty. Many projects use both.

DimensionPrimary trackComparison track
Complete cycleFrom defining problems to undergoing review
Step oneDefine problems and assumptions
Step twoReview literature and select methods
Step threeDerive or model and compute
Step fourVerify results and boundaries
Step fiveDocument and present findings
Step sixUndergo review and revise

Core responsibilities

  1. 01

    Mentor others on mathematical techniques.

  2. 02

    Maintain knowledge in the field by reading professional journals, talking with other mathematicians, and attending professional conferences.

  3. 03

    Develop new principles and new relationships between existing mathematical principles to advance mathematical science.

  4. 04

    Disseminate research by writing reports, publishing papers, or presenting at professional conferences.

What are the qualification boundaries?

In the U.S., this occupation is in Job Zone 5. Most positions require graduate education (about 45% doctoral and 25% master's). Check title and project reviews with institutions and responsible authorities. This page does not determine individual eligibility.

  1. Boundaries

    Verify jurisdiction, institution, and academic standards

  2. Jurisdiction

    Follow rules in the relevant country or region

  3. Institution

    Confirm with universities, research institutions, and responsible authorities

  4. Training

    Confirm education, research, and teaching experience

  5. Responsibility

    Confirmed by the researcher and project lead

  6. Evidence

    Retain verifiable records

Work reality

Work reality: O*NET 15-2021.00 respondents report mainly indoor work with written and team communication. Research requires sustained derivation, computation, and review; verify working hours by institution and project. Key judgments concern clarity of problems and assumptions, method and model applicability, sufficient proof or verification, controllable numerical error, and compliance with citation and research output boundaries.

O*NET structured fields

Research inputs
Research inputs
Research inputs: Occupational identity, tasks, work activities, work context, skills and knowledge, interest code, education, and Job Zone for O*NET 15-2021.00. Wage, employment, and industry data on that page cover this occupation. Research findings are not used to infer mainland Chinese qualifications, pay, or employment definitions.

Mathematicians' objects of work and AI boundaries Tools for symbolic computation and theorem-proving assistance, numerical computation and optimization, literature and data searches, and paper and handout drafting can assist with calculation verification, idea exploration, and material organization. Researchers remain responsible for problem definition, assumption selection, proof and modeling judgments, conclusion reliability, and academic standards. Tools must not replace rigorous reasoning and review.

Career direction comparison

Comparison with adjacent occupations

The following compares mathematicians and adjacent roles by actual objects of work, core deliverables, and method boundaries. Qualifications are not interchangeable based on titles.

Comparison with adjacent occupations
Career directionCore work and outputKey differencesBest-fit choice
Financial Quantitative AnalystsFinancial quantitative analysts focus on financial market models and trading strategies; mathematicians focus on mathematical theory and applied research, with different objectives and deliverables.Focus on financial modeling and trading strategies.Compare if you prefer financial quantitative work.
Computer and Information Research ScientistsComputer and information research scientists focus on frontier computing and information science research; mathematicians focus on mathematical theory and applications, with different methodological emphases.Focus on frontier computing and information research.Compare if you prefer computing research.
Operations Research AnalystsOperations research analysts focus on resource and decision optimization; mathematicians focus on mathematical theory and broader applications, with different sources of problems.Focus on resource and decision optimization.Compare if you prefer optimization applications.
Statisticians and BiostatisticiansStatisticians and biostatisticians focus on data inference and biomedical applications; mathematicians focus on mathematical theory and applications in other fields, with different objectives and verification methods.Focus on data inference and biomedical applications.Compare if you prefer statistical inference.

How to choose

First use a verifiable research or modeling explanation to check assumptions, method applicability, and conclusion boundaries, then assess a direction. Do not assign a single ranking based on income or personality.

Next, turn the comparison into a small-scale research exercise that can be stopped.

AI impact on this career

Boundaries of AI impact on mathematicians

Tools for symbolic computation and theorem-proving assistance, numerical computation and optimization, literature and data searches, and paper and handout drafting can assist with calculation verification, idea exploration, and material organization. Researchers remain responsible for problem definition, assumption selection, proof and modeling judgments, conclusion reliability, and academic standards. Tools must not replace rigorous reasoning and review. AI Impact Score 58% is an editorial estimate. Model impact index for three equally weighted representative core tasks; not a measured share of working time or a job-loss probability. floor((50+75+50)/3+0.5)=58; fermatmind-ai-impact-v1; 2026-10-18.The three tasks, ratings and sources are listed in the FermatMind method source record on this page.

AI assessment method and sources

FermatMind AI Impact Score estimation method v1

FermatMind / 费马测试 · 2026-10-18

  • Model impact index for three equally weighted representative core tasks; not a measured share of working time or a job-loss probability. floor((50+75+50)/3+0.5)=58; fermatmind-ai-impact-v1; 2026-10-18.
  • Anchors: 0 negligible; 25 preparation or record assistance; 50 substantial assistance; 75 most information output can be formed by AI with review; 100 almost all of the selected task can be done by AI and independently verified.
  • Derive and verify mathematical conclusions — 50/100; weight=1/3
  • Model and perform numerical computation — 75/100; weight=1/3
  • Apply mathematics to practical problems — 50/100; weight=1/3

Editorial ratings use the existing occupational responsibilities and AI task analysis on this page. They are not adoption rates, measured effects, job-disappearance or personal-unemployment probabilities, or official scores from Stanford or another external institution. No measured time weights or future robot deployment are assumed.

Technical capabilities

Describe only specific steps tools can assist with; do not treat demonstrations as institutional adoption.

Adoption conditions

Insufficient evidence: No mathematician-specific adoption rate or controlled effect evidence was obtained. Check inputs, tool reliability, and review responsibility before trials.

Job outcomes

Without occupational-level job outcome research, do not infer layoff timing, staffing changes, or hiring probabilities.

Task impact matrix

Comparison of AI impact across career tasks
Work areaTaskCurrent changeHuman control point
Derivation and proofDerive and verify mathematical conclusionsEditorial recommendation: Tools may assist symbolic computation, numerical experiments, and drafting. Current materials do not demonstrate widespread institutional adoption of these systems.Check assumptions and reasoning.
Modeling and computationModel and perform numerical computationMay assist computationFollow assumptions, method applicability, and academic standards, retaining human judgment and records.
Applications and collaborationApply mathematics to practical problemsNo automated decisionsConfirmed by the researcher and application stakeholder.
Writing and reviewDocument findings and undergo reviewCannot replace judgments involving responsibilityConfirmed by the researcher and reviewers.

Why does employment evidence reach different conclusions?

Occupational sources describe actual tasks and objects of work: problems and assumptions, models and proofs, computation and data, and papers and reports. Tool capabilities, institutional adoption, and job outcomes are recorded separately. Wage or employment data are not substitutes for AI evidence.

O*NET 15-2021.00 Mathematicians (Summary)

View source ↗
Subject
Identity and scope of 15-2021.00
Finding
Supports identity and task boundaries
Limitation
U.S. scope for this occupation; excludes Chinese data.

O*NET 15-2021.00 Tasks

View source ↗
Subject
Tasks for 15-2021.00
Finding
Supports task breakdowns, not AI adoption rates
Limitation
Does not replace local law or academic standards.

Generative AI and Jobs: A Refined Global Index of Occupational Exposure (ILO Working Paper 140, 2025)

View source ↗
Subject
Global Occupational Task Exposure Index (ILO 2025)
Finding
Background for task exposure analysis
Limitation
Not mathematician-specific adoption rates, experimental effects, or causal evidence of layoffs.

Why does impact differ by work direction?

Editorial analysis: Try tools for symbolic computation, numerical experiments, and drafting, comparing accuracy, verifiability, and cost for each. Existing occupational task and exposure research cannot establish which stage will see adoption first or greater employment impact.

Derivation and proof

AI primarily changesEditorial recommendation: Try proof assistance, while still checking reasoning step by step.

Human responsibility remainsResponsibility for assumptions and reasoning

Modeling and computation

AI primarily changesEditorial recommendation: Try computation and optimization, while still checking errors and models.

Human responsibility remainsModel applicability judgments

Accountability must remain intact when AI enters professional workflows

Human control points: Define problems and assumptions, judge method and model applicability, verify conclusions and errors, take responsibility for findings and citation standards, and follow academic ethics and funding agreements.

  1. Check inputs

    Confirm problems, assumptions, and standards.

  2. Limit tasks

    Conduct research only within academic and ethical boundaries.

  3. Human review

    Review reasoning, calculations, and citations.

  4. Document and escalate

    Promptly report standards and ethical issues.

Professional risks of using AI

Assumption and method applicability

May affect whether conclusions hold.

Control: Check applicability conditions and review.

Proof and conclusion reliability

May lead to incorrect conclusions.

Control: Reason rigorously and undergo review.

Computational and numerical errors

May affect result credibility.

Control: Control errors and cross-validate.

How should different audiences respond?

Entrants

Practice derivation, computation, and writing.

Current practitioners

Retain reasoning and computation records.

Managers

Set standards and review requirements first.

Common questions about AI and career development

Will AI replace mathematicians?

Task exposure, adoption, and job outcomes measure different things and cannot establish a date for job cuts.

Source: O*NET 15-2021.00 TasksOriginal source ↗

Which stage will automation affect first?

Currently, only assistable symbolic computation, numerical experiments, and drafting are described. Humans retain responsibility for assumption selection, proof judgments, and conclusions.

Source: O*NET 15-2021.00 TasksOriginal source ↗

Does this page have adoption-rate data?

There is no occupational-level adoption-rate source, so this page does not provide such figures.

Source: O*NET 15-2021.00 Mathematicians (Summary)Original source ↗

First conduct an offline exercise with an entirely fictional problem context and data. Write modeling assumptions, a solution or proof approach, and verification methods without relying on unverified automated computation or proofs to replace rigorous reasoning.

Chinese mainland salary reference

This page did not obtain a national occupational median wage for mathematicians in mainland China. National Bureau of Statistics 2025 Tables 2 and 5 publish average annual wages by broad industry category. Research institution positions may reference scientific research and technical services, while university positions fall under education or other industries. A uniform industry classification cannot be inferred from the occupation's name.

Scope note: National Bureau of Statistics 2025 average annual wages for persons employed in scientific research and technical services were RMB 182,064 per year in urban non-private units and RMB 83,560 per year in urban private units; the national total for urban non-private units was RMB 129,441 per year. All are industry-level figures, not this occupation's median or starting pay.

Official statistical scope

Category boundaries: Ministry of Human Resources and Social Security enterprise compensation survey information is classified by occupation and position and can be used to look up research and technical service pay. This page did not obtain an entry precisely corresponding to O*NET 15-2021.00 and does not use it as this occupation's national median. Refer to releases from the responsible authorities.

National · Scientific research and technical services (non-private)

Only average annual wages are published

National · Scientific research and technical services (private)

Only average annual wages are published

Precise wages for this occupation

Insufficient evidence: Not separately listed

Scenario reference

Insufficient evidence: A national occupational median wage for mathematicians in mainland China was not obtained. The figures below are 2025 average annual industry wages in urban units, not starting pay or individual take-home income.

This scenario explains statistical scope only; it is not an individual pay quote.
Role scopeReference valueExplanation
Not separately reportedAppointment by employer and positionNational Bureau of Statistics wage data are classified by industry and registration type, not education levels for this occupation.
Not separately reportedAppointment by employer and positionNational Bureau of Statistics wage data are classified by industry and registration type, not education levels for this occupation.
Not separately reportedAppointment by employer and positionNational Bureau of Statistics wage data are classified by industry and registration type, not education levels for this occupation.
Not separately reportedAppointment by employer and positionNational Bureau of Statistics wage data are classified by industry and registration type, not education levels for this occupation.

Influencing factors

Scientific research and technical services (non-private)

RMB 182,064 per year

Scientific research and technical services (private)

RMB 83,560 per year

National total for urban non-private units

RMB 129,441 per year

Boundaries for AI and pay assessments

Insufficient evidence: A national occupational median wage, statistical period, and definition for mathematicians in mainland China were not obtained. This page displays only labeled industry proxies and does not fill gaps with proxy or U.S. numbers.

U.S. salary reference

U.S. pay and employment scope for mathematicians (occupation 15-2021.00)

U.S. national mathematicians (15-2021): BLS 2025 median annual wage of $126,710, with the 10th percentile at $69,240 and the 90th at $195,190. These are not starting pay or seniority levels. BLS Table 1.2 reports about 2,200 U.S. jobs in 2025, +0.6% projected growth in 2025–2035, and about 100 average annual openings. The overall projection is not attributed to AI and does not equal current job postings or individual opportunities. BLS Table 1.2 for 15-2021: about 2,200 U.S. jobs in 2025, about 2,200 projected in 2035 (+0.6%), and about 100 annual openings. The table rounds counts independently and openings include replacements.

BLS 2025 wages listed separately for mathematicians: The median and 10th and 90th percentiles describe the national wage distribution, not minimum or maximum wages, seniority levels, or take-home income.

BLS 2025 wages listed separately for mathematicians: The median and 10th and 90th percentiles describe the national wage distribution, not minimum or maximum wages, seniority levels, or take-home income.
PercentileAnnual referenceHourly equivalentExplanation
10th percentile$69,240 per year—About 10% of U.S. mathematician wages in 2025 were below this value; it is not minimum pay or beginner income.
Median (50%)$126,710 per year$60.92 per hour2025 U.S. national median annual wage (occupation 15-2021.00).
90th percentile$195,190 per year—About 10% of U.S. mathematician wages in 2025 were above this value; it is not maximum pay or senior income.

Pay · Median:$126,710 per year (2025);O*NET 15-2021.00 National Wages (BLS OEWS display)

Pay · Hourly median:$60.92 per hour (2025);O*NET 15-2021.00 National Wages (BLS OEWS display)

Pay · 10th percentile:$69,240 per year (2025);BLS OOH Mathematicians: 2025 Wages and 2025–2035 Outlook

Pay · 90th percentile:$195,190 per year (2025);BLS OOH Mathematicians: 2025 Wages and 2025–2035 Outlook

Pay · Statistical scope:Mathematicians (15-2021), U.S. national;O*NET 15-2021.00 National Wages (BLS OEWS display)

Percentiles are not seniority: About 10% of wages are below the 10th percentile and about 10% above the 90th. Institutions, regions, and tasks differ; beginner or senior income cannot be inferred from percentiles.

Wages

2025 national median and percentiles, not starting pay, seniority, or take-home income.

Employment

The 2025 base measures jobs, not vacancies.

Outlook

BLS Table 1.2 for 15-2021: about 2,200 U.S. jobs in 2025, about 2,200 projected in 2035 (+0.6%), and about 100 annual openings. The table rounds counts independently and openings include replacements.

Industry

The table below shows 2025 median annual wages for mathematicians in specific industries, not averages for all employees in those industries.

Major employers of mathematicians in 2025 include the federal government and educational institutions. The table below uses specific industry median annual wages from the mathematician subtable in BLS Pay, separately from industry employment shares.

Major employers of mathematicians in 2025 include the federal government and educational institutions. The table below uses specific industry median annual wages from the mathematician subtable in BLS Pay, separately from industry employment shares.
Industry2025 median annual wageHow to interpret
Aerospace product and parts manufacturing$185,520 per year2025 BLS OOH Pay industry subtable listed separately for mathematicians; national median annual wage for mathematicians in this industry, not an individual pay quote.
Professional, scientific, and technical services$140,660 per year2025 BLS OOH Pay industry subtable listed separately for mathematicians; national median annual wage for mathematicians in this industry, not an individual pay quote.
Federal government$136,990 per year2025 BLS OOH Pay industry subtable listed separately for mathematicians; national median annual wage for mathematicians in this industry, not an individual pay quote.
State colleges, universities, and professional schools$65,510 per year2025 BLS OOH Pay industry subtable listed separately for mathematicians; national median annual wage for mathematicians in this industry, not an individual pay quote.

BLS Table 1.2 for 15-2021: about 2,200 U.S. jobs in 2025, about 2,200 projected in 2035 (+0.6%), and about 100 annual openings. The table rounds counts independently and openings include replacements.

Qualifications

O*NET respondents report education of 45% doctorates, 25% master's degrees, and 10% bachelor's degrees. BLS states that a master's or higher is typical, with some positions accepting bachelor's degrees. Respondent percentages are not shares of job postings.

Experience

Editorial recommendation: Demonstrate experience through verifiable research, modeling, and teaching outputs. Current sources do not compare the relative hiring validity of papers and certificates.

Region

Strongly affected by the distribution of government and research institutions; do not extrapolate across regions.

Employment outlook and job openings

Fit signals

Fit signals: Start with observable work behaviors

To assess fit, consider whether you can sustain abstract derivation and computation, reason rigorously and treat uncertainty honestly, model practical problems mathematically, read literature and write, and continue producing work in a competitive environment. Tools such as RIASEC only generate hypotheses and cannot replace actual task experience.

Positive signals: Positive signals: Abstract thinking and rigorous reasoning, sustained focus, mathematical modeling of practical problems, honesty about uncertainty, and continued reading and writing.

Reasons for caution: Reasons for caution: Difficulty sustaining abstract derivation, a tendency to skip verification or overstate conclusions, unwillingness to write and undergo review, or difficulty handling competition and long periods without results.

Trial action: Trial action: With an entirely fictional problem context and data, practice developing modeling assumptions, a solution approach, and verification methods, then ask someone to review them against your records. When comparing a mathematical modeling post with a statistics post, ask who defines the constraints and what counts as a model failure. For a fictional dispatch problem, submit two capacity constraints and a small counterexample that makes the proposed schedule impossible. Ask whether the team expects a proof, a simulation or a recommendation, and who checks that it applies outside the worked example.

Do not let one test decide your career

RIASEC

Work sample

Check through work samples rather than substituting labels for judgment.

The O*NET interest code is IC (Investigative–Conventional): an inclination toward research and analysis and organizing symbolic information according to procedures. These labels describe activity preferences only and cannot determine competence or qualifications.

  • Observable tasks
  • Process records

Do not use it this way: Hiring, qualifications, or performance cannot be inferred from this.

Start a career interest exercise

Big Five personality

Work sample

Check through work samples rather than substituting labels for judgment.

General personality measures such as the Big Five provide no job-specific evidence for this occupation and support self-reflection only. Continue checking through actual derivation and proof, modeling and computation, and writing and presentation experience.

  • Observable tasks
  • Process records

Do not use it this way: Hiring, qualifications, or performance cannot be inferred from this.

Start a career interest exercise

MBTI

Work sample

Check through work samples rather than substituting labels for judgment.

Type tools such as MBTI and the Enneagram cannot determine competence or qualifications and support self-reflection only.

  • Observable tasks
  • Process records

Do not use it this way: Hiring, qualifications, or performance cannot be inferred from this.

Start a career interest exercise

Enneagram

Work sample

Check through work samples rather than substituting labels for judgment.

Continue checking against actual tasks, boundaries, and verifiable deliverables.

  • Observable tasks
  • Process records

Do not use it this way: Hiring, qualifications, or performance cannot be inferred from this.

Start a career interest exercise

Cognitive ability

Work sample

Check through work samples rather than substituting labels for judgment.

Continue checking against actual tasks, boundaries, and verifiable deliverables.

  • Observable tasks
  • Process records

Do not use it this way: Hiring, qualifications, or performance cannot be inferred from this.

Start a career interest exercise

Emotional ability

Work sample

Check through work samples rather than substituting labels for judgment.

Continue checking against actual tasks, boundaries, and verifiable deliverables.

  • Observable tasks
  • Process records

Do not use it this way: Hiring, qualifications, or performance cannot be inferred from this.

Start a career interest exercise

Which mathematics research direction might suit you better?

Derivation and proof

Fits better when: Can derive rigorously and verify conclusions.

Watch for: Must check assumptions and boundaries.

Modeling and computation

Fits better when: Can model and perform numerical computation.

Watch for: Must check methods and errors.

Abstraction and application

Fits better when: Can apply mathematics to practical problems.

Watch for: Must check application context.

Literature and writing

Fits better when: Can read literature and document findings.

Watch for: Must check standards and citations.

Presentations and communication

Fits better when: Can give academic presentations and exchange ideas.

Watch for: Must check audience and scope.

Mentoring and collaboration

Fits better when: Can guide others and collaborate in research.

Watch for: Must handle task allocation and authorship.

Career fit questions

How can I identify the work focus in a job posting?

Theoretical research defines a problem, develops a proof, and discusses its implications. Applied research clarifies assumptions with collaborators, compares models, and explains where they work. Compare proof and research outputs with model delivery.

What should I ask in an interview?

If an assumption looks unreliable, can I reformulate the problem or should I provide a result conditional on that assumption?

Can you sustain abstract derivation?

Use this page's work sample to check focus and reasoning.

Can you rigorously verify conclusions?

Check reasoning and errors against the method.

Can you apply mathematics to practical problems?

Assess modeling and application skills.

What would you do if assumptions or methods seemed questionable?

Check first and indicate uncertainty.

Can you sustain reading and writing?

Assess literature and writing skills.

Can you take responsibility for conclusions and standards?

Assess responsibility and academic ethics awareness.

How can your work be reviewed?

Enable others to review it through reasoning and findings records.

What exercise should you try next?

Complete an exercise with an entirely fictional problem context and data.

Risks and boundaries

Work risks and boundaries for mathematicians

Address risks through triggers, consequences, controls, and escalation boundaries.

Editorial risk analysis: Uses O*NET mathematician tasks and work context to develop scenarios for checking assumption applicability, proofs, errors, citations, and output pressure. It is not evidence of incident rates or statutory liability. Certification directories do not support academic integrity conclusions about generative tools.

01

Assumption and method applicability risks

Typical scenario:Insufficient conditions for assumptions or method applicability.

Roles most affected:Research and modeling positions

Possible consequence:May affect whether conclusions hold.

How to reduce risk:Check applicability conditions and review.

Evidence:O*NET 15-2021.00 Tasks

02

Proof and conclusion reliability risks

Typical scenario:Insufficiently rigorous reasoning or conclusions.

Roles most affected:Research positions

Possible consequence:May lead to incorrect conclusions.

How to reduce risk:Reason rigorously and undergo review.

Evidence:O*NET 15-2021.00 Tasks

03

Computational and numerical error risks

Typical scenario:Uncontrolled numerical computation or simulation errors.

Roles most affected:Computation and analysis positions

Possible consequence:May affect result credibility.

How to reduce risk:Control errors and cross-validate.

Evidence:O*NET 15-2021.00 Work Context

04

Academic standards and citation risks

Typical scenario:Citations or findings fail to meet academic standards.

Roles most affected:Research and publication positions

Possible consequence:May violate academic integrity.

How to reduce risk:Follow standards and check citations.

Evidence:O*NET 15-2021.00 Mathematicians (Summary)

05

Competition and output pressure risks

Typical scenario:Competition and sustained output pressure.

Roles most affected:Research positions

Possible consequence:May affect career stability and health.

How to reduce risk:Manage research pace and seek collaboration.

Evidence:O*NET 15-2021.00 Work Context

06

Tool use beyond academic boundaries

Typical scenario:Using generative tools without reviewing reasoning or citations.

Roles most affected:Research positions

Possible consequence:May violate academic integrity.

How to reduce risk:Limit uses and review reasoning and citations.

Evidence:O*NET 15-2021.00 Tasks

Entry pathways

Entry pathways and capability stages

In the U.S., this occupation is in Job Zone 5 (extensive preparation needed). Most positions require graduate education; respondents report about 45% doctorates, 25% master's degrees, and 10% bachelor's degrees. Extensive related experience is usually needed. Verify specific requirements with institutions and employers.

Mathematics research

01

Research Assistant

Typical scope:Entry stage: Under supervision, practice literature reading and problem definition, basic derivation and numerical computation, modeling and programming, academic writing, and presentations. Learn academic standards and retain process records.

Readiness evidence:Evidence to show: Use a verifiable research or modeling explanation, including problem and assumptions, methods and calculations, conclusions and limitations, to demonstrate rigorous reasoning and honest documentation.

Credential boundary:Verify education and project requirements.

Next moves:Next step: Check the target institution's research direction and requirements, findings and citation standards, and project and funding conditions. Verify modeling, computation, and writing skills in supervised real research.

02

Mathematicians

Typical scope:Conduct independent research and undergo review.

Readiness evidence:Retain reasoning and findings records.

Credential boundary:Verify institutional and title requirements.

Next moves:Progression to greater responsibility: With experience, lead independent projects and directions, mentor and collaborate in teams, develop a specialty, or move into applied research, data analysis, or teaching.

03

Senior Researcher

Typical scope:Lead research directions and projects.

Readiness evidence:Retain direction and project records.

Credential boundary:Verify title requirements.

Next moves:Build project and mentoring evidence.

04

Research Lead

Typical scope:Lead teams and funding.

Readiness evidence:Retain team and funding records.

Credential boundary:Verify institutional requirements.

Next moves:Build governance and strategy evidence.

Transition to applied mathematics and data

01

Applied Work / Data Assistant

Typical scope:Participate in applied work and data analysis under supervision.

Readiness evidence:Retain analysis and application records.

Credential boundary:Verify employer requirements.

Next moves:Build data and modeling foundations.

02

Applied Mathematician or Data Scientist

Typical scope:Independently handle modeling and analysis.

Readiness evidence:Retain model and analysis records.

Credential boundary:Verify employer requirements.

Next moves:Build business and engineering evidence.

03

Senior Modeling / Data Lead

Typical scope:Lead complex models and solutions.

Readiness evidence:Retain complex model records.

Credential boundary:Verify employer requirements.

Next moves:Build evidence of methodological systems.

04

Data Science Lead

Typical scope:Lead teams and methodological standards.

Readiness evidence:Retain team and standards records.

Credential boundary:Verify employer requirements.

Next moves:Build evidence of management responsibility.

Transition to teaching and communication

01

Teaching Assistant / Lecturer

Typical scope:Participate in teaching and course support under supervision.

Readiness evidence:Retain teaching and course records.

Credential boundary:Verify institutional requirements.

Next moves:Build teaching foundations.

02

Lecturer / Assistant Professor

Typical scope:Independently teach courses and conduct research.

Readiness evidence:Retain course and findings records.

Credential boundary:Verify institutional and title requirements.

Next moves:Build evidence of teaching and results.

03

Associate Professor

Typical scope:Lead curricula or research directions.

Readiness evidence:Retain system and direction records.

Credential boundary:Verify title requirements.

Next moves:Build governance and impact evidence.

04

Professor / Discipline Lead

Typical scope:Lead disciplinary direction and teams.

Readiness evidence:Retain discipline and team records.

Credential boundary:Verify title requirements.

Next moves:Build strategy and governance evidence.

Capability ladder: what demonstrates readiness?

  1. 01

    Check inputs

    Confirm problems, assumptions, and standards.

  2. 02

    Process control

    Derive, model, and verify according to methods.

  3. 03

    Results review

    Create verifiable reasoning and findings records.

  4. 04

    Expanded responsibility

    Expand project and application responsibilities within authorization.

Market signals

Market signals and transition directions

BLS Table 1.2 for 15-2021: about 2,200 U.S. jobs in 2025, about 2,200 projected in 2035 (+0.6%), and about 100 annual openings. The table rounds counts independently and openings include replacements.

Three sources, three different measures

U.S. national

2025–2035

U.S. national · Display scope: Mathematicians (15-2021.00)

BLS Table 1.2 for 15-2021: about 2,200 U.S. jobs in 2025, about 2,200 projected in 2035 (+0.6%), and about 100 annual openings. The table rounds counts independently and openings include replacements.

Employment size and growth

BLS Table 1.2 for 15-2021: about 2,200 U.S. jobs in 2025, about 2,200 projected in 2035 (+0.6%), and about 100 annual openings. The table rounds counts independently and openings include replacements.

Does not establish:U.S. occupational data. The projection does not measure current recruitment or individual outcomes.

U.S. national

2025–2035

U.S. national · Display scope: Mathematicians (15-2021.00)

About 100 per year

Average annual job openings

Annual openings include replacements.

Does not establish:U.S. occupational data. The projection does not measure current recruitment or individual outcomes.

U.S. national

2025

U.S. national · Display scope: Mathematicians (15-2021.00)

$126,710 per year

Median annual wage

BLS 2025 OEWS national median, not starting pay or take-home income.

Does not establish:National median listed separately for mathematicians; excludes individual take-home income.

Where can your current capabilities transfer?

These are capability relationships, not automatic promotions or credential waivers.

Financial Quantitative Analysts

Editorial assessment: Evidence to build

Transfers:Transferable capabilities: Abstract reasoning and proof, mathematical modeling and numerical computation, programming and algorithms, data analysis, academic writing and presentations, cross-disciplinary applications, and collaboration.

Build next:Gaps to address: Applied knowledge and programming skills in the target field, statistics and machine learning methods, or experience and evidence of responsibility for transitions into operations optimization, data science, and quantitative analysis.

View career profile

Computer and Information Research Scientists

Editorial assessment: Evidence to build

Transfers:Mathematics and computation foundations.

Build next:Build computing and information research expertise.

View career profile

Operations Research Analysts

Editorial assessment: Evidence to build

Transfers:Modeling and optimization foundations.

Build next:Build operations research and decision methods.

View career profile

Statisticians

Editorial assessment: Evidence to build

Transfers:Mathematics and statistics foundations.

Build next:Build statistical inference and data methods.

View career profile

Biostatisticians

Editorial assessment: Evidence to build

Transfers:Mathematics and statistics foundations.

Build next:Build biomedical statistics expertise.

View career profile

Data Scientists

Editorial assessment: Evidence to build

Transfers:Mathematics and modeling foundations.

Build next:Build data engineering and business experience.

View career profile

FAQs and boundaries

What are daily work and working hours like?

Mathematicians on this page refers to O*NET 15-2021.00: professionals who research fundamental mathematics or applications of mathematical techniques to science, management, and other fields and solve problems using mathematical methods. Work includes fundamental mathematics research, developing new relationships among mathematical principles, formulating hypotheses and deriving consequences, numerical analysis and computation, mathematical modeling, applying mathematical techniques to practical problems, developing computational methods, writing papers, giving presentations, and instructing others in mathematical methods. Work is mainly in offices and computing environments. This is distinct from financial quantitative analysts, computer and information research scientists, operations research analysts, statisticians, and biostatisticians.

How does this occupation differ from adjacent occupations?

Financial quantitative analysts focus on financial modeling and trading; computer and information research scientists on computing and information research; operations research analysts on resource and decision optimization; statisticians and biostatisticians on data inference and biomedical applications. Mathematicians focus on mathematical theory and broader applications. Compare objects, methods, and deliverables to distinguish them.

What signals suggest this occupation might suit me?

Observable signals: Abstract thinking and rigorous reasoning, sustained focus, mathematical modeling of practical problems, honesty about uncertainty, and continued reading and writing. The O*NET interest code IC describes activity preferences only and cannot determine competence or qualifications.

How will AI affect this occupation?

Tools for symbolic computation and theorem-proving assistance, numerical computation and optimization, literature and data searches, and paper and handout drafting can assist calculation verification, idea exploration, and material organization. Researchers remain responsible for problem definition, assumption selection, proof and modeling judgments, conclusion reliability, and academic standards. Tools must not replace rigorous reasoning and review. This page presents tool uses only as task hypotheses, not evidence of widespread institutional adoption.

Which judgments must remain human responsibilities?

Human responsibilities include defining problems and assumptions, judging method and model applicability, verifying conclusions and errors, taking responsibility for findings and citation standards, and following academic ethics and funding agreements.

Source: FermatMind|Global task level|FermatMind task rubric v2026.09; reviewed 2026-09-25|internal task rubric
What pay evidence is available for mainland China?

The National Bureau of Statistics publishes average annual wages for persons employed in urban units by broad industry category, without separately listing this occupation, so a safely mapped national occupational median cannot be established. The page displays only labeled industry proxies: in 2025, scientific research and technical services recorded RMB 182,064 per year in urban non-private units and RMB 83,560 per year in urban private units; the national total for urban non-private units was RMB 129,441 per year. None equals this occupation's median or starting pay. Ministry of Human Resources and Social Security enterprise compensation survey information is a lookup resource.

What is the U.S. pay evidence?

The O*NET page displays a 2025 BLS OEWS median annual wage of $126,710 (about $60.92 per hour) for occupation 15-2021.00. That page does not display percentile wages. These are U.S. national statistics, not individual take-home income or starting pay.

What are the entry and training pathways?

In the U.S., this occupation is in Job Zone 5. Most positions require graduate education (about 45% doctoral, 25% master's, and 10% bachelor's) and usually extensive related experience. Check title and project reviews with institutions and responsible authorities. In mainland China, there is no uniform occupational qualification precisely corresponding to this O*NET occupation; check employer and industry requirements.

What are the main risks and control boundaries?

Main risks involve assumption and method applicability, proof and conclusion reliability, computational and numerical errors, academic standards and citations, and competition and output pressure. Controls include checking assumptions and applicability conditions, rigorous reasoning and verification, error control, following academic standards, and managing output pace. Check specific rules with institutions and applicable standards.

How should market signals be understood?

BLS Table 1.2: U.S. occupation 15-2021 had about 2,200 jobs in 2025 and is projected to have about 2,200 in 2035 (+0.6%), with about 100 openings per year. Counts are rounded independently; projected openings include replacements and do not measure current hiring. This occupation is very small and its figures are sensitive to fluctuations. These are occupational statistics, not an individual's hiring probability.

Is this occupation worth exploring further?

If you can sustain abstract derivation and rigorous verification, try a low-cost exercise with an entirely fictional problem context and data. Moving into operations optimization, data science, or quantitative analysis requires applied expertise and experience evidence.

What are this page's usage boundaries?

Evidence for China, the U.S., and specific institutions is presented separately: conclusions about pay, employment, qualifications or credentials, safety, and AI must not be combined across scopes.; The mathematician page provides occupational assessment and research practice information and does not replace local laws, academic standards, ethical requirements, or other professional advice.; 'Insufficient evidence' means only that safely corresponding data were not established through a bounded search; proxy data or memory are not used to fabricate numbers.

Sources

FermatMind research workflow

2025 BLS OEWS wage data; accessed 2026-09-25

O*NET 15-2021.00 National Wages (BLS OEWS display)

  • O*NET 15-2021.00 Mathematicians (Summary)O*NET OnLine, U.S. Department of Labor|United States|O*NET 2026 occupational profile; accessed 2026-09-25|official occupation summary
    View original source:O*NET 15-2021.00 Mathematicians (Summary)
    • Provides occupational identity, official description, interest code, Job Zone, skills, and knowledge for 15-2021.00.
    • Wage and employment data on that page cover this occupation (15-2021.00).

    Limitations:Used only for 15-2021.00 and the U.S. market; does not establish a mapping to China.

  • O*NET 15-2021.00 TasksO*NET OnLine, U.S. Department of Labor|United States|O*NET 2026 occupational profile; accessed 2026-09-25|official occupation task list
    View original source:O*NET 15-2021.00 Tasks
    • Provides tasks including fundamental mathematics research, developing new relationships among mathematical principles, formulating hypotheses and deriving consequences, numerical analysis and computation, developing mathematical models for analysis or simulation, applying mathematical techniques to business, engineering, and scientific problems, developing computational methods, writing reports and papers, giving academic presentations, and instructing others in mathematical methods.
    • The task sample does not mean every position performs every task.

    Limitations:The task sample does not mean every position performs every task; it does not establish a mapping to China.

  • O*NET 15-2021.00 Work ContextO*NET OnLine, U.S. Department of Labor|United States|O*NET 2026 occupational profile; accessed 2026-09-25|official occupation profile
    View original source:O*NET 15-2021.00 Work Context
    • Describes mainly controlled indoor work, communication through documents and meetings, high decision-making freedom, an emphasis on precision, strong competition, and workweeks usually exceeding 40 hours.
    • U.S. work context does not replace Chinese law or employer procedures.

    Limitations:U.S. work context does not replace Chinese law or employer procedures.

  • O*NET 15-2021.00 National Wages (BLS OEWS display)O*NET OnLine using U.S. Bureau of Labor Statistics data|United States|2025 BLS OEWS wage data; accessed 2026-09-25|O*NET display of BLS OEWS wage data
    View original source:O*NET 15-2021.00 National Wages (BLS OEWS display)
    • 2025 BLS OEWS median wage; the page covers occupation 15-2021.00.
    • Percentile wages are not displayed on this page; inferred values are not substituted.

    Limitations:Percentile wages are not displayed on this page; the median does not represent starting pay or individual take-home income.

  • U.S. BLS Table 1.2: Occupational projectionsU.S. Bureau of Labor Statistics|United States|2025–2035|official employment data
    View original source:U.S. BLS Table 1.2: Occupational projections
    • BLS Table 1.2: U.S. occupation 15-2021 had about 2,200 jobs in 2025 and is projected to have about 2,200 in 2035 (+0.6%), with about 100 openings per year. Counts are rounded independently; projected openings include replacements and do not measure current hiring.
    • Projections are not hiring guarantees.

    Limitations:Projections are not hiring guarantees; 'decline' does not mean individual opportunities will necessarily decrease.

  • Industries: 15-2021.00O*NET OnLine, U.S. Department of Labor|United States|O*NET industry data current 2026; accessed 2026-09-25|official industry distribution
    View original source:Industries: 15-2021.00
    • The main industries on that page for 2024 are Government and Educational Services.
    • Industry distribution signals employer settings, not projected growth or demand.

    Limitations:Industry distribution signals employer settings, not projected growth or demand.

  • O*NET 15-2021.00 Education and Career PreparationO*NET OnLine, U.S. Department of Labor|United States|O*NET 2026 occupational profile; accessed 2026-09-25|official occupation profile
    View original source:O*NET 15-2021.00 Education and Career Preparation
    • Job Zone 5 (extensive preparation needed): Most positions require graduate education. Respondents report about 45% requiring a doctorate, 25% a master's degree, and 10% a bachelor's degree. Extensive related experience is usually needed.
    • The education distribution does not constitute mandatory requirements for every employer.

    Limitations:The education distribution does not constitute mandatory requirements for every employer; it does not establish a mapping to China.

  • Licenses & Certifications: 15-2021.00O*NET OnLine, U.S. Department of Labor|United States|current page; accessed 2026-09-25|official license and certification directory
    View original source:Licenses & Certifications: 15-2021.00
    • O*NET search tools for U.S. state licenses and national certifications.
    • Certificates and training do not equal authorization to practice or employer hiring requirements.

    Limitations:Certification and training do not equal a scope of practice or employment; check the applicable jurisdiction.

  • market_signalInternational Labour Organization|Global task research|ILO Working Paper 140, 2025; accessed 2026-09-25|official research report
    View original source:market_signal
    • Assesses occupational exposure to generative AI at the task level for task-level discussion.
    • The exposure index does not equal adoption rates, job outcomes, or hiring probabilities.

    Limitations:Task exposure does not equal implemented automation, adoption rates, or layoffs; individual outcomes are not inferred.

  • Average Annual Wages of Persons Employed in Urban Units in 2025National Bureau of Statistics|National scope, mainland China|2025 annual wages; published 2026-05-15; accessed 2026-09-25|official industry wage statistics
    View original source:Average Annual Wages of Persons Employed in Urban Units in 2025
    • Provides average annual wages by broad industry category for urban non-private and private units, including scientific research and technical services, to check public Chinese wage definitions.
    • An industry average is not an occupational median and excludes individual businesses and freelancers.

    Limitations:An industry average is not an occupational median and excludes individual businesses and freelancers; no national wage distribution precisely corresponding to this occupation was established.

  • Enterprise Compensation Survey Information (2024)Ministry of Human Resources and Social Security|Urban enterprises in mainland China|2024 annual data; published 2025-06; accessed 2026-09-25|official enterprise wage survey
    View original source:Enterprise Compensation Survey Information (2024)
    • Ministry of Human Resources and Social Security enterprise compensation survey information, a pay lookup resource classified by occupation and position.
    • Covers urban enterprises, excluding government agencies and public institutions; this page did not obtain an entry precisely corresponding to 15-2021.00.

    Limitations:Covers urban enterprises, excluding government agencies and public institutions; this page did not obtain an entry precisely corresponding to this occupation and does not use it as a national median.

  • FermatMind Mathematicians Task Impact ScaleFermatMind|Global task level|FermatMind task rubric v2026.09; reviewed 2026-09-25|internal task rubric
    • Organizes tasks that can be assisted and human control points for task-level discussion of AI and fit.
    • This editorial scale is not evidence of adoption rates, layoffs, regulations, or licensing.

    Limitations:This editorial scale is not evidence of adoption rates, layoffs, regulations, or licensing.

  • BLS OOH Mathematicians: 2025 Wages and 2025–2035 OutlookU.S. Bureau of Labor Statistics|U.S. national|May 2025 wages; 2025—2035 projections; accessed 2026-10-11|official occupation wage and employment projections
    View original source:BLS OOH Mathematicians: 2025 Wages and 2025–2035 Outlook
    • The Pay subtable separately lists 2025 mathematician wage percentiles and industry medians; these are not combined mathematician and statistician wages.
    • Job Outlook separates combined-group projections from the mathematician subcategory. Openings are annual averages; overall projections are not attributed to AI.

    Limitations:This page uses only separately listed mathematician wages. The combined group's 10% and 2,000 average annual openings must not be treated as mathematician data.

  • Three representative core tasks of this occupationFermatMind / 费马测试|2026-10-18|Editorial model estimation method
    View original source:Three representative core tasks of this occupation
    • Model impact index for three equally weighted representative core tasks; not a measured share of working time or a job-loss probability. floor((50+75+50)/3+0.5)=58; fermatmind-ai-impact-v1; 2026-10-18.
    • Anchors: 0 negligible; 25 preparation or record assistance; 50 substantial assistance; 75 most information output can be formed by AI with review; 100 almost all of the selected task can be done by AI and independently verified.
    • Derive and verify mathematical conclusions — 50/100; weight=1/3; display.ai_impact_table.task_rows.0[0]
    • Model and perform numerical computation — 75/100; weight=1/3; display.ai_impact_table.task_rows.1[0]
    • Apply mathematics to practical problems — 50/100; weight=1/3; display.ai_impact_table.task_rows.2[0]

    Limitations:Editorial ratings use the existing occupational responsibilities and AI task analysis on this page. They are not adoption rates, measured effects, job-disappearance or personal-unemployment probabilities, or official scores from Stanford or another external institution. No measured time weights or future robot deployment are assumed.

  • U.S. BLS Table 1.2: Occupational projectionsU.S. Bureau of Labor Statistics|United States|2025–2035
    View original source:U.S. BLS Table 1.2: Occupational projections
    • BLS Table 1.2: U.S. occupation 15-2021 had about 2,200 jobs in 2025 and is projected to have about 2,200 in 2035 (+0.6%), with about 100 openings per year. Counts are rounded independently; projected openings include replacements and do not measure current hiring.
  • O*NET® 31.0 Core Task StatementsU.S. Department of Labor, Employment and Training Administration|United States|O*NET 31.0|Primary document
    View original source:O*NET® 31.0 Core Task Statements

    Limitations:Verbatim Core Task Statements excerpted from the O*NET® 31.0 Database, used under CC BY 4.0. U.S. occupational scope.

  • O*NET® database license · CC BY 4.0U.S. Department of Labor, Employment and Training Administration|United States|O*NET 31.0|Primary document
    View original source:O*NET® database license · CC BY 4.0

    Limitations:Verbatim Core Task Statements excerpted from the O*NET® 31.0 Database, used under CC BY 4.0. U.S. occupational scope.

U.S. occupational data. The projection does not measure current recruitment or individual outcomes.

Mathematician: Proofs, Models and Assumptions | FermatMind