Preparing content…

Physicists

SOC 19-2012 · O*NET 19-2012.00

Physicists · models, experiments, and evidenceReproducible evidence · clear model boundariesWage gaps marked · experiment safety first
50%AI Impact ScoreFermatMind / 费马测试

Physicists study physical phenomena using mathematical models, experiments or computation. They document measurement conditions and uncertainty so colleagues can evaluate the findings.

Model impact index for three equally weighted representative core tasks; not a measured share of working time or a job-loss probability. floor((50+25+75)/3+0.5)=50; fermatmind-ai-impact-v1; 2026-10-18.

172,250 dollars per yearUnited States 19-2012 median annual wageU.S. Department of Labor, O*NET
About 1,300 per yearUnited States 19-2012 average annual openings (including replacements)U.S. Bureau of Labor Statistics
Source: U.S. Bureau of Labor Statistics|United States|2025 base period; 2025—2035 overall projection (not attributed to AI)|Average annual position openings (including growth and replacements; not attributed to AI); Editorial conversion: BLS 19-2012 standalone row, 1.3 thousand per year; annual openings include replacements, and the projected net increase is about 1,700 people.
No data availableMainland China salary reference

Quick judgment

What kind of people are suited to physicists?

Paper check | Compare two fictional measurement sheets. Flag a change in units and identify which metadata to check before comparing values.

Result review: ask a physics teacher or researcher to check variable definitions, units, error propagation, and the scope of the conclusion; one paper exercise cannot prove research ability.

Who is more likely to fit Physicists?

Safe experience: using public data or fully fictional pendulum measurement data, fit the relationship between period and pendulum length and record assumptions, calculations, residuals, and possible bias; do not touch high-voltage, radiation, laser, or other experimental equipment.

When should you be cautious about choosing Physicists?

Direct judgment: if you are willing to describe phenomena mathematically, patiently troubleshoot experimental or simulation errors, and revise hypotheses when results do not match expectations, you can continue exploring physics research. Positions may lean toward theory, experiment, computing, or application.

How can a small experiment test your fit for Physicists?

Observable signals: able to write a problem as a testable hypothesis, state units, measurement conditions, and uncertainty; separate data, model output, and interpretation; and be willing to let others recompute.

Career profile

Physicists study physical phenomena using mathematical models, experiments or computation. They document measurement conditions and uncertainty so colleagues can evaluate the findings. When reviewing a job, ask who checks the work or records and which tasks require separate authorization.

Career quick answers

What categories can physicist work be divided into?

Physicists explain physical phenomena with testable models and evidence; theoretical derivation, experimental measurement, computational simulation, and technical application may be carried out by different teams.

Theoretical and computational research:

Key judgment: pose a hypothesis from a physics question, choose an experiment or simulation that can discriminate between explanations, control variables, check dimensions and calibration, analyze errors, and review conclusions.

  1. 01Pose the question
  2. 02build a model
  3. 03design measurement or simulation
  4. 04collect data
  5. 05check units and errors
  6. 06recompute
  7. 07publish or apply results

Experiment and measurement:

Quality standard: hypotheses and variable definitions are clear, raw data and code are traceable, instrument calibration or simulation parameters are reviewable, and uncertainty and the scope of conclusions are explicit.

  1. 01Research question
  2. 02literature and hypothesis
  3. 03method design
  4. 04data or simulation
  5. 05error analysis
  6. 06peer review
  7. 07paper, report, or technology transfer
Applied physics R&DInspectable deliverables: a research plan, parameter and data dictionary, analysis code, error budget, charts, and a paper or technical report; peers should be able to recompute them.

Where are the boundaries between physicists and adjacent positions?

Physicists focus on physical mechanisms and evidence interpretation, astronomers on celestial research, materials scientists on material properties, and engineers on design and engineering delivery; methods can be shared, while position requirements and final outputs require separate verification. Complete loop: when findings and model predictions disagree, recheck the instrument, data processing, and assumptions and record revisions, rather than selecting only the results that support expectations.

DimensionPrimary trackComparison track
Core purposePose testable physics questions and explain phenomena and test predictions with theory, experiment, or simulation.The core evidence is reproducible data, models, errors, and peer challenge; fluent explanation or a single image cannot replace measurement verification. Research project cycles, experimental equipment access time, academic publication, and funding deadlines affect the pace; judge by the specific position rather than the occupational group.
Main objects and basisExperimental and applied physics: design measurements, calibrate instruments, analyze data, and verify application conditions with engineering teams.Theoretical and computational physics: derive models, solve equations, and run simulations, comparing predictions with observable results.
Core processDefine assumptions, methods, and variables, obtain data or run simulations, analyze errors, and accept peer review.Adjacent boundary: astronomers focus on celestial bodies in the universe, materials scientists on material properties, and engineers on technical implementation; research methods can be shared but final outputs differ. Task reminder: the O*NET tasks support computation, experiment, simulation, papers, and teaching; the BLS 19-2012 wages and employment are the standalone United States physicist statistics.
Main outputsInspectable deliverables: reproducible analysis of public or fictional data, experimental design, and error notes; such work cannot replace degree or position requirements.Investment targets: advanced mathematics, core physics courses, experimental and computational methods, and research projects; research and university positions usually require a doctorate.
Responsible partyResearchers are responsible for the methods, data, code, and conclusions they claim; project leaders and peers review according to their actual responsibilities.Risk loop: control dimensions, calibration, sample selection, code versions, simulation boundaries, and research integrity; hazardous equipment is operated by qualified personnel per institutional rules. The United States 19-2012 median annual wage in 2025 was 172,250 dollars; the occupational percentiles are not an individual starting salary or Chinese position income.
Authority and approvalVerify data permissions, equipment training, and institutional authorization; a physics degree does not automatically grant the right to operate hazardous equipment.Different directions differ in pressure from experiment scheduling, publication, funding, and product cycles; verify the specific position's work outputs and evaluation criteria. Tools and materials: mathematical models, public or authorized data, computational software, measurement instruments, and experimental records; tool suggestions should be checked against physics and the data.
Conclusion boundariesQuality standard: consistent dimensions, calibrated equipment, reproducible code, reasonable statistical and error handling, and conclusions that can withstand peer challenge.Collaboration chain: mentors or project leaders, laboratory technicians, engineers, data analysts, and peer reviewers; specify parameters, versions, and handover.

Core responsibilities

  1. 01

    Analyze data from research conducted to detect and measure physical phenomena.

  2. 02

    Report experimental results by writing papers for scientific journals or by presenting information at scientific conferences.

  3. 03

    Perform complex calculations as part of the analysis and evaluation of data, using computers.

  4. 04

    Describe and express observations and conclusions in mathematical terms.

What must be verified item by item before entering?

Qualification boundary: BLS states that research and academic positions usually require a doctorate and some federal entry positions may accept a bachelor's degree; specific employer requirements require separate verification.

  1. Target position degree

    Verify the target announcement; research and university positions usually require a doctorate, and federal entry-level physics positions usually require a bachelor's degree in physics.

    O*NET 19-2012 tasks include research data analysis, complex calculations, instruments, theory, and papers; not all physicists can be equated with laboratory operators. Inputs: research questions, models, instruments, or public data; outputs: verified explanations, predictions, papers, software models, or application plans.

  2. Physics direction and foundation

    Verify mathematics, core physics, programming, and laboratory courses and match them to a theory, experiment, or application topic.

    O*NET supports physics tasks and IRC, BLS supports the 19-2012 wages, overall employment, and education path; the NBS scientific research industry and broad professional technical position averages serve only as Chinese background, not a physics occupational median. Risks: measurement error, model extrapolation, code errors, experiment safety, and selective reporting can distort conclusions; review and records are required.

  3. Research and project experience

    Verify mentor projects, original contributions, and reproducible methods; the number of papers does not replace quality.

    BLS Table 1.2 United States 19-2012 physicists 23,200 in 2025 and 24,800 in 2035, overall growth 7.2%; the original table's ten-year net increase is 1,700, and the displayed endpoint difference of 1,600 cannot replace the unrounded net increase. Average annual openings 1,300 including growth and replacements, not a net increase or immediate hiring; the overall projection is not attributed to AI and does not guarantee opportunities in China or a specific research direction. Environment: universities, government laboratories, and industrial R&D differ in funding, project deadlines, instrument resources, and publication requirements.

  4. Data and equipment authorization

    Verify data permissions and hazardous equipment training, approval, protection, and operating authority.

    Materials: public data and fictional measurements can be used for experience; real experiments, confidential data, and hazardous equipment are constrained by institutional rules. Collaboration: explain physical assumptions, parameters, and uncertainty in interdisciplinary teams so others can recompute and judge the applicable range of the plan.

  5. Employer and funding conditions

    Verify the school, institute, or company duties, contract term, funding source, and security clearance conditions.

    Review: check raw data, processing flow, dimensions, calibration, code versions, error propagation, and independent repeat results. Classification: 19-2012 is physicists; astronomers, materials scientists, nuclear engineers, and technicians each have different occupation codes.

  6. Outputs and evaluation methods

    Verify papers, simulations, measurements, or application metrics and the feedback method, and assess the ongoing training investment.

    Path: build a foundation in mathematics, physics, programming, and experiments during the bachelor's degree; research and university positions usually require a doctorate, and specific entry positions depend on the posting. Decision: first choose a theory, experiment, computing, or application setting, then verify course and research training, position degree requirements, experiment safety, and showable results.

Work reality

Workplace
Universities, research institutes, government laboratories, and R&D companies.
Research objects
Physics problems in matter, energy, particles, fields, light, celestial bodies, or materials.
Core tools
Mathematics, programming, measurement instruments, simulation software, and statistical methods.
Collaboration partners
Physics peers, engineers, technicians, and interdisciplinary researchers.
Decision responsibility
State assumptions, measurement conditions, the model's applicable range, and uncertainty clearly.
Quality requirements
Data traceable, calculations reproducible, conclusions not beyond the evidence.
Time rhythm
Experiment scheduling, project milestones, and paper review create periodic pressure.
Safety boundary
Experimental equipment and data are operated and managed per institutional rules.
Position differences
Theoretical, experimental, computational, and applied research differ in tools and outputs.

O*NET structured fields

19-2012.00BLS Occupational Outlook HandbookNational Bureau of Statistics 2025 urban unit wagesBLS employment projections table 1.2 (2025–35)BLS physicist education path
19-2012.00
O*NET 19-2012 lists tasks such as analyzing research data on physical phenomena, complex calculations, designing experimental equipment, building theories, running simulations, and publishing results; the specific task frequency varies by position.
BLS Occupational Outlook Handbook
Environment and pace: theoretical positions emphasize derivation and programming, experimental positions emphasize instruments and measurement, and applied positions emphasize prototypes and performance constraints; the same title does not guarantee the same laboratory time. Responsibility boundary: physicists are responsible for research interpretation or physical solutions; engineers are responsible for product constraints and engineering delivery, and technicians operate equipment according to training and authorization.
National Bureau of Statistics 2025 urban unit wages
China's scientific research industry and broad professional technical position pre-tax annual averages, not the physicist occupational median; specific university, institute, and company positions require separate verification.
BLS employment projections table 1.2 (2025–35)
BLS Table 1.2 physicists 19-2012: 23.2 thousand in 2025 to 24.8 thousand in 2035, growth of 7.2%, an original-table net increase of 1.7 thousand; 1.3 thousand average annual openings including replacements, not attributed to AI.
BLS physicist education path
BLS states that research and university positions usually require a doctorate in physics, astronomy, or a related field; United States federal entry-level physics positions usually require a bachelor's degree in physics. Specific announcements and institutional equipment training authorization require separate verification.

Career direction comparison

How do physicists differ from adjacent research and engineering positions?

Physicists center on physics questions, models, measurement, and verifiable explanation. Astronomers, materials scientists, mathematicians, computer and information research scientists, and nuclear engineers share methods, but their research objects and final deliverables differ.

How do physicists differ from adjacent research and engineering positions?
Career directionCore work and outputKey differencesBest-fit choice
PhysicistsStudy physical phenomena, build models, and verify them with experiments or simulations.Explanations, predictions, or physical methods must state the evidence and applicable range.Suits people willing to work with mathematics, measurement, and uncertainty over the long term.
AstronomersStudy stars, planets, and cosmic phenomena and analyze observational data.The research object is celestial bodies; observation conditions and data sources differ.Suits people interested in astronomical observation and cosmic questions.
Materials scientistsStudy material structure, properties, and processing relationships.Material performance and characterization are the core, and physics principles are an important tool.Suits people interested in material synthesis, characterization, and performance validation.
MathematiciansStudy mathematical structures, proofs, or applied mathematical models.Mathematical proofs and methods can stand alone as results and do not necessarily verify the physical world experimentally.Suits people who prefer abstract reasoning and rigorous proof.
Nuclear engineersDesign and evaluate nuclear-related systems, equipment, and safety plans.Delivers engineering systems that meet technical specifications; safety authorization requires separate verification.Suits people who apply physics principles to constrained engineering systems.
Computational physics directionBuild numerical models, algorithms, and simulations and compare predictions with observations.An internal direction within physics positions; correct code does not guarantee correct physical assumptions.Suits people interested in mathematical modeling, programming, and reproducible verification.

How to choose

Sharing mathematics, programming, or experimental equipment does not mean the positions are the same; compare research questions, testable results, engineering constraints, and employer requirements.

First choose theoretical, experimental, computational, or applied research, then verify courses, degree, research experience, equipment safety, and position deliverables; experience in adjacent occupations does not automatically equal physics research qualification.

AI impact on this career

How does AI change the specific tasks of physicists?

An approved tool can organize a draft handoff from fictional inputs. Check original values, revisions and open items, then refer the draft to the person authorized to review it. Example: Compare two fictional measurement sheets. Flag a change in units and identify which metadata to check before comparing values.

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+25+75)/3+0.5)=50; 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.
  • Method design|Choose theoretical derivation, experiment, or simulation|Specify variables, boundary conditions, and controls. — 50/100; weight=1/3
  • Data acquisition|Operate equipment or obtain authorized data according to training and rules|Keep calibration, parameters, and raw records. — 25/100; weight=1/3
  • Analysis and interpretation|Check dimensions, errors, and alternative explanations|Do not equate a model fit with proof of causation. — 75/100; 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.

Task exposure

It means a type of task may be assisted by AI, not an adoption rate or employment decline.

Model capability

A demonstration can generate text or code and cannot prove physical correctness, experiment safety, or research reproducibility.

Occupational projection

The BLS employment projection reflects multiple factors and does not separate out the net AI effect.

Task impact matrix

Comparison of AI impact across career tasks
Work areaTaskCurrent changeHuman control point
Research backgroundLiterature retrievalCan summarize paper abstracts and candidate literatureVerify the original text, method, period, and applicability to the question.
Hypothesis designResearch question breakdownCan list variables, alternative models, and items to be measuredResearchers define the testable predictions and exclusionary evidence.
Computational implementationAnalysis code draftCan write function and test draftsVerify the algorithm, dimensions, numerical stability, and benchmark examples.
Data processingData cleaning and plottingCan flag outliers and generate chartsVerify raw records, calibration, point-removal rules, and error bars.
Numerical researchSimulation parameter explorationCan propose parameter ranges and comparison plansVerify boundary conditions, convergence, model assumptions, and independent recomputation.
Experimental collaborationExperiment plan organizationCan organize materials, steps, and record templatesQualified personnel verify equipment, safety training, controls, and measurement uncertainty.
Research expressionPaper and report draftsCan organize method and result textAuthors verify citations, formulas, results, and limitations and bear research integrity responsibility.

Why does employment evidence reach different conclusions?

O*NET gives occupational tasks, BLS gives employment and wages, and ILO and NIST give cross-occupation task exposure and a governance framework; these materials do not measure the physicist AI replacement rate.

O*NET 19-2012

View source ↗
Subject
United States physicist tasks
Finding
Lists research data, computation, experiment, theory, and paper tasks
Limitation
Does not measure AI product effects, adoption rates, or replacement rates

ILO generative AI study

View source ↗
Subject
Cross-occupation task exposure
Finding
Suggests potential change at the task level
Limitation
Cannot infer the number of standalone physicist positions

NIST AI risk framework

View source ↗
Subject
AI governance principles
Finding
Supports validity, transparency, and risk management
Limitation
Does not validate the correctness of a specific physics research tool

BLS 19-2012 projection

View source ↗
Subject
2025—2035 United States occupational employment
Finding
2025—2035 United States 19-2012 overall growth 7.2%, average annual openings 1,300 including replacements, and a ten-year net increase of 1,700, not attributed to AI.
Limitation
The projection does not separate out the causal AI contribution

An editorial summary based on O*NET tasks

View source ↗
Subject
Physics research work chain
Finding
Retrieval and code may be trialed with assistance, and conclusions must be reviewed physically and against the data
Limitation
The editorial summary is not a product evaluation or a position replacement rate

Why does impact differ by work direction?

Separate material that can be organized automatically from evidence that requires physics judgment: tool output is a hypothesis awaiting testing, and research conclusions must undergo measurement, derivation, or independent recomputation.

Literature and problem definition

AI primarily changesCan summarize public research and candidate explanations

Human responsibility remainsResearchers verify the source, published version, physical assumptions, and testable questions.

Programming and data processing

AI primarily changesCan draft code, cleaning scripts, and plots

Human responsibility remainsResearchers verify the data dictionary, units, outliers, algorithms, and versions.

Models, simulation, and experiment

AI primarily changesCan suggest parameters or simulation plans

Human responsibility remainsResearchers verify boundary conditions, calibration, uncertainty, and independent repetition.

Accountability must remain intact when AI enters professional workflows

Trial within authorized data and tools, keeping raw inputs, code versions, and manual review records; high-risk experiment safety still follows institutional rules.

  1. Define the problem

    State the research question, data permissions, tool purpose, and conclusions that cannot be decided automatically.

  2. Check the input

    Keep literature versions, raw data, units, instruments, or simulation parameters.

  3. Validate the output

    Test with benchmark examples, limiting cases, an error budget, and independent recomputation.

  4. Manual review

    A person with the relevant physics ability reviews the model, interpretation, and paper claims.

  5. Records and improvement

    Keep versions and causes of error; when experimental equipment is involved, escalate according to institutional safety rules.

Professional risks of using AI

Dimensional or formula errors

It would make apparently reasonable numbers lose physical meaning

Control: Do a dimensional check, limiting-case test, and independent derivation.

Data processing bias

Removing points, interpolating, or converting units may change the conclusion

Control: Keep the raw data and processing scripts and recompute.

Misuse of simulation boundaries

A fitted result may go beyond the model's valid domain

Control: Publish the parameters, boundary conditions, and uncertainty.

Fabricated citations or results

It would contaminate the literature review and the paper

Control: Verify the original text, DOI, figures, and data sources item by item.

Experiment safety misdirection

Tool text cannot authorize operating hazardous equipment

Control: Operate and review only by trained personnel according to institutional rules.

How should different audiences respond?

New to the field or changing direction

Practice dimensions, calculus, experimental error, and reproducible code; use public data for small recomputations, and do not treat model answers as physical truth.

Already in a research position

Set a verification checklist for literature, code, and simulation assistance, and check data sources, parameters, model boundaries, and results.

Research team leaders

Specify data authorization, versions, review, experiment safety, and error escalation responsibility, and record the tool's applicable range.

Hiring or study decision makers

Look at the physics methods, research outputs, and degree required by the position; do not use generalized AI predictions to replace current hiring information.

Common questions about AI and career development

Can AI produce research conclusions for physicists?

Conclusions cannot be drawn from model output alone; physical assumptions, data, errors, and reproducibility must be verified.

Source: O*NET tasks and editorial summaryOriginal source ↗

Will AI replace experimental measurement?

Tools can assist design and analysis, but calibration, real measurement, equipment safety, and abnormal review still require qualified personnel.

Source: O*NET tasks and NIST governanceOriginal source ↗

Which physics tasks are more suitable for trying AI first?

Public literature organization, non-sensitive code drafts, and chart templates are easier to set up for manual review; use public or authorized data first.

Source: O*NET physics tasks and editorial summaryOriginal source ↗

Will physicist positions decrease by 2035?

BLS Table 1.2 United States 19-2012 physicists 23,200 in 2025 and 24,800 in 2035, overall growth 7.2%; the original table's ten-year net increase is 1,700, and the displayed endpoint difference of 1,600 cannot replace the unrounded net increase. Average annual openings 1,300 including growth and replacements, not a net increase or immediate hiring; the overall projection is not attributed to AI and does not guarantee opportunities in China or a specific research direction.

Source: BLS standalone occupational projectionOriginal source ↗

Can AI-generated physics formulas be used directly?

No. Check dimensions, limiting cases, derivation conditions, and consistency with experiments or trusted worked examples.

Source: NIST governance and editorial adviceOriginal source ↗

In which year will physicists disappear?

There is no reliable 19-2012 disappearance year or specific replacement rate; cross-occupation AI exposure cannot infer this result.

Source: ILO cross-occupation studyOriginal source ↗

What should I learn first now?

First master mathematics, physics experiments, errors, programming, and research integrity, then learn how to verify AI-assisted output.

Source: BLS education path and O*NET tasksOriginal source ↗

First learn physical modeling, error analysis, and reproducible experiments, then use AI to improve retrieval or code drafts; always check the output against the original evidence.

Chinese mainland salary reference

'Physicists' on this page corresponds to the United States Physicists statistical row, not a name for an independent practice license in China.

The United States code 19-2012 is used only for United States physicist statistics; Chinese hiring must verify the research direction, education path, and job duties.

Scientific research industry and broad professional technical position wage reference

United States 19-2012 is the standalone physicist statistical row; Chinese position titles, degrees, and institutional hiring requirements must be verified against the original job posting.

National official statistics

No comparable occupation-specific value for 19-2012

Specific research institutions

Current written quotes must be verified

United States physicist statistics

Cannot be converted into a Chinese wage

Which job directions do physicists usually start from under different education backgrounds?

Insufficient evidence: no national occupation-specific wage median in mainland China directly corresponding to United States 19-2012 was verified.

Education background, common job directions, and verification suggestions (editorial compilation, not wage statistics)
Education backgroundCommon job directionsNotes and verification suggestions
Verify the doctorate and research outputsUniversity research positionsBLS states that United States research and university physics positions usually require a doctorate; Chinese positions must be verified against the original job posting.
Verify the position degree and project requirementsGovernment or public research institutionsDifferent research institutions have different hiring requirements, which cannot be presumed from United States federal entry position rules.
Verify the physics direction and project skillsCorporate R&D positionsPositions may require experimental, computational, or device experience; the same title does not mean the same degree.
Research usually requires a doctorate, and some federal entry positions may accept a bachelor's degreeUnited States physicistsThe BLS occupational outlook explains the United States path; the specific job posting takes priority.

The NBS scientific research and technical services wages cover all positions in the industry, and the broad professional technical position group includes scientific research personnel; neither lists national occupational wages for Chinese physicists separately, and specific university, institute, and company positions require separate verification.

Which industries and directions use this occupation?

Universities and research institutes

Position demand depends on the discipline direction, funding, and specific hiring and cannot be converted from the United States projection.

Corporate R&D

Verify the technical topic, product stage, project experience, and current hiring.

Public laboratories

Verify the job posting, equipment authorization, and institutional safety requirements.

Will AI affect the wages of physicists?

AI may assist with code and literature organization; this time there is no evidence to quantify wage changes or position replacement rates for Chinese physicists.

United States salary reference

How much do United States physicists earn?

The 2025 United States 19-2012 physicist national annual wage 10 / 25 / 50 / 75 / 90 percentiles are 82,110, 111,720, 172,250, 222,760, and 274,110 dollars; not new entrant income or a Chinese quote. BLS Table 1.2 United States 19-2012 physicists 23,200 in 2025 and 24,800 in 2035, overall growth 7.2%; the original table's ten-year net increase is 1,700, and the displayed endpoint difference of 1,600 cannot replace the unrounded net increase. Average annual openings 1,300 including growth and replacements, not a net increase or immediate hiring; the overall projection is not attributed to AI and does not guarantee opportunities in China or a specific research direction. The United States 19-2012 physicist 2025 median annual wage was 172,250 dollars.

United States wage distribution

United States wage distribution
Wage statistical positionAnnual wage referenceConversion boundariesHow to understand this position
10th percentileBelow 82,110 dollars per yearOfficial annual wage, not individual monthly incomeThe 2025 United States 19-2012 national 10th percentile is 82,110 dollars per year, with about 10% of salaried workers at or below this value; it is not a new entrant starting wage, an entry-level grade, or individual after-tax pay. The percentiles are not grouped by seniority, research direction, or city, and the specific contract requires separate verification.
Median (50%)172,250 dollars per yearOfficial annual wage, not individual take-home payThe median is the 50% percentile, the midpoint of the distribution: half of salaried employees earn more than it per year and half earn less. This page uses the officially published median and does not substitute an average. It mixes people from different regions, industries, and experience levels, so it cannot be treated as an entry-level wage or used to derive the income of a specific position or individual; when comparing across regions, also note the statistical period and whether bonuses are included.
90th percentileAbove 274,110 dollars per yearOfficial annual wage, not individual take-home payThe 2025 United States 19-2012 national 90th percentile is 274,110 dollars per year, with about 10% of salaried workers at or above this value. The percentile does not prove a higher grade, greater responsibility, or bonus composition, does not guarantee individual promotion income, and cannot be converted into a Chinese wage.

Wages · 10th percentile:82,110 dollars per year;2025 10th percentile: about 10% of salaried physicists were below this value; not a new entrant starting wage.;O*NET national wage percentiles

Wages · 25th percentile:111,720 dollars per year;2025 25th percentile: a national distribution position, not grouped by research direction or degree.;O*NET national wage percentiles

Wages · median:172,250 dollars per year;2025 median: the midpoint of the 19-2012 physicist occupation wage distribution, not individual take-home pay.;BLS occupational outlook or projections table

Wages · 75th percentile:222,760 dollars per year;2025 75th percentile: a national occupation position, not equal to a position quote.;O*NET national wage percentiles

Wages · 90th percentile:274,110 dollars per year;2025 90th percentile: about 10% of salaried workers were above this value, not a promotion promise.;O*NET national wage percentiles

How should this set of United States wage figures be understood?

Is the median an entry-level wage?

No. 172,250 dollars is the 2025 United States 19-2012 physicist occupational wage median, not grouped by new entrants.

Can hourly wages be used to estimate an individual annual salary?

An individual hourly wage, after-tax income, or research funding contract cannot be derived from the median annual wage; the specific position must be checked.

Which statistical program do these figures come from?

Wage percentiles come from the BLS 2025 occupational wage data displayed through O*NET; employment comes from the BLS 2025—2035 occupational projection.

Which conditions affect actual income?

Direction, location, degree, research experience, institution, and contract all affect individual wages; national percentiles provide background only.

Are 19-2012 wages comparable across industries?

Are 19-2012 wages comparable across industries?
IndustryMedian annual wage / statistical statusHow to understand this value
State colleges, universities, and professional schools84,600 dollars per yearBLS 2025 median annual industry wage for physicists in state colleges, universities, and professional schools, not a quote for all universities or new entrants.
Scientific research and development services176,170 dollars per yearBLS 2025 median annual wage for physicists in the scientific research and development services industry, not for all corporate R&D or individual funding contracts.
Federal government (excluding postal service)146,900 dollars per yearBLS 2025 median annual wage for physicists in the federal government (excluding postal service), not a single laboratory pay grade or a Chinese public position.
Outpatient medical services239,360 dollars per yearBLS 2025 median annual wage for physicists in the outpatient medical services industry, not a clinical physician wage; specific medical physics qualifications require separate verification.

What factors affect United States physicist wages?

Research direction and employer

Universities, government laboratories, and corporate R&D differ in position goals and pay structures; national percentiles are not grouped by direction.

Location and contract

City, working hours, funding method, benefits, and employment contract affect individual income; national percentiles are not a monthly salary promise.

Degree and project experience

The doctorate, research experience, and technical direction affect the position scope, but wage percentiles cannot directly explain causation or seniority.

United States 19-2012 overall employment projection (not attributed to AI)

Employment · average annual position openings

1,300 per year

BLS Table 1.2 average annual openings 1,300 including growth and replacements; a ten-year net increase of 1,700 people, not annual openings, and not attributed to AI.;BLS Table 1.2

Source: U.S. Bureau of Labor Statistics|United States|2025 base period; 2025—2035 overall projection (not attributed to AI)|Average annual position openings (including growth and replacements; not attributed to AI); Editorial conversion: BLS 19-2012 standalone row, 1.3 thousand per year; annual openings include replacements, and the projected net increase is about 1,700 people.

Career fit guide

What interests and work styles suit physicists?

First consider whether you enjoy repeated verification in mathematical reasoning, experiments, or simulations and are willing to record uncertainty. The O*NET 19-2012 interest code is IRC (Investigative, Realistic, Conventional); interests only help exploration and cannot prove research ability or hiring qualification.

You may be better suited to: Willing to turn vague physical phenomena into testable questions and check dimensions, errors, and alternative explanations.

These situations require caution: If you are unwilling to handle long-cycle failed experiments, repeated recomputation, or rigorous records, observe a real research project first.

Do an entry practice exercise: Use public or fictional data to complete a small modeling and error analysis and seek feedback from a physics teacher or researcher.

Do not let one test decide your career

Holland interests

Occupational tasks and editorial exploration advice; not a hiring threshold

How do I verify that Holland interests match physics work?

The O*NET 19-2012 interest code is IRC: Investigative, Realistic, and Conventional; observe whether you enjoy explaining regularities, working with instruments or code, and documenting accurately.

  • Enjoys mathematics and experiments
  • Willing to verify data and models

Do not use it this way: An interest code is only an occupational description and cannot presume research ability.

Test my Holland career interests

Big Five

Occupational tasks and editorial exploration advice; not a hiring threshold

How do I verify that Big Five traits match physics work?

Focus on openness, conscientiousness, and work habits under pressure; trait scores cannot replace a physics foundation or team projects.

  • Willing to explore complex problems
  • Can keep records of errors over time

Do not use it this way: Personality scales are not a hiring threshold.

See my Big Five personality traits

Myers-Briggs Type Indicator (MBTI)

Occupational tasks and editorial exploration advice; not a hiring threshold

Can MBTI determine physics research competence?

MBTI can be used to discuss study and collaboration preferences and cannot judge physics research ability by four-letter type; still verify courses, projects, and recomputation performance. Use real courses, projects, and recomputation performance to verify mathematics, programming, experimental, and writing ability.

  • Can reproduce simplified results
  • Can explain sources of error

Do not use it this way: A self-rating score is not equal to position ability.

Understand my MBTI preferences

Enneagram

Occupational tasks and editorial exploration advice; not a hiring threshold

How is the Enneagram used for career exploration?

The Enneagram can help reflect on reactions to failure, challenge, and long-cycle pressure; an unverified type does not predict research outcomes or hiring. Compare preferences for long-term research, publication, technical application, pay, and location against the target institution's position conditions.

  • Accepts the research cycle
  • Can face uncertain conclusions

Do not use it this way: Preferences do not predict employment outcomes.

Explore my Enneagram patterns

Intelligence quotient (IQ) and reasoning

Occupational tasks and editorial exploration advice; not a hiring threshold

Can an IQ score replace physics ability evidence?

An unverified IQ hiring cutoff cannot be set; look at whether one can derive models, identify dimensions and errors, recompute, and explain results. Compare sustained investment in derivation, experiment, and computation projects and seek specific mentor feedback.

  • Can pose testable hypotheses
  • Willing to redo failed analyses

Do not use it this way: One experience cannot represent long-term fit.

Understand my intelligence and reasoning performance

Emotional intelligence (EI)

Occupational tasks and editorial exploration advice; not a hiring threshold

How is emotional intelligence observed in teamwork?

Observe whether one can communicate clearly under challenge and failure, take feedback, and hand over accurately; an EQ score cannot replace research quality and safety training. Cross-check with public-data projects and target position requirements before deciding on a course or application direction.

  • Work can be recomputed
  • Position requirements have been verified

Do not use it this way: No assessment can replace a job posting.

Understand my emotional intelligence abilities

Which physics or adjacent research direction are you more likely suited to?

Astronomer direction

Fits better when: Enjoys celestial observation and cosmic questions

Watch for: Research objects and data sources differ; physics research experience cannot be treated as position qualification

Explore related career: Astronomers

Materials scientist direction

Fits better when: Enjoys material structure, characterization, and performance

Watch for: Material development deliverables differ from theoretical physics research

Explore related career: Materials scientists

Mathematician direction

Fits better when: Enjoys abstract proofs and mathematical structures

Watch for: Mathematical proofs can stand alone as results and do not necessarily require physical measurement

Explore related career: Mathematicians

Nuclear engineer direction

Fits better when: Enjoys nuclear system design constrained by codes and standards

Watch for: Engineering safety and authorization require separate verification; a physics background does not automatically grant authority

Explore related career: Nuclear engineers

Computer and information research scientist direction

Fits better when: Enjoys research on algorithms and computing systems

Watch for: Algorithm research outputs differ from physical interpretation

Explore related career: Computer and information research scientists

Postsecondary physics teacher direction

Fits better when: Enjoys teaching and academic research

Watch for: University position degree, teaching, and research requirements require separate verification

Explore related career: Postsecondary physics teachers

Career fit questions

Do physicists need to be geniuses?

Positions need solid mathematics and physics training, reproducible methods, and continuous correction; being a genius is not a verifiable hiring indicator.

Can I do physics if I prefer programming?

Computational physics is one direction, but the code must correspond to the physical assumptions, boundary conditions, and data.

Can I still do physics if I dislike experiments?

Theoretical and computational directions can reduce experimental operations but still need observable results to test claims.

Does enjoying hands-on work mean I am suited?

Interest in experiments helps, and safety training, dimensions, calibration, and error analysis are also needed.

What is the O*NET interest code?

19-2012 interest code IRC: Investigative, Realistic, Conventional; it only describes preferences for occupational activities.

Can assessments prove that I am suited?

No. Courses, research projects, mentor feedback, and position requirements are more explanatory than a single score.

Must physicists have a doctorate?

BLS states that research and university physics positions usually require a doctorate; some federal entry positions may accept a bachelor's degree, and specific positions require separate verification.

What if I care more about product application?

Applied physics, materials, or engineering R&D positions can be considered, comparing final deliverables, engineering constraints, and qualifications.

How do I distinguish interest from ability?

Interest shows whether you are willing to do it long term, and ability shows whether you can derive, measure, recompute, and explain uncertainty.

How can I get a low-risk experience this week?

Use public or fictional pendulum data to fit a model and write the units, residuals, errors, and limitations, and ask a teacher for feedback.

Work pressure, risks, and occupational boundaries

What are the main work risks for physicists?

Focus on preventing dimensional or model errors, measurement and calibration bias, irreproducible code or data processing, experimental equipment safety, and extrapolation of conclusions. Each requires traceable records and peer review.

O*NET supports the physicist research tasks and BLS supports the occupation path and employment scope; the specific control actions are editorial advice, not incident rates or statutory safety procedures.

01

Dimensions and model assumptions

Typical scenario:Formula values look reasonable but the units are inconsistent or the boundary conditions do not hold.

Roles most affected:Physics researchers, collaborating engineers

Possible consequence:Predictions or explanations are distorted and subsequent experiments and application directions are misled.

How to reduce risk:State the units and boundaries clearly and review with limiting cases, independent derivation, and benchmark examples.

Evidence:O*NET physicist tasks · NIST AI risk management framework

02

Measurement and calibration bias

Typical scenario:Instrument drift, sample selection, or missing records shift the results.

Roles most affected:Experimental physicists, technicians

Possible consequence:Papers or plans are built on incomparable data.

How to reduce risk:Keep calibration, raw data, environmental conditions, and an error budget; repeat the measurement when abnormal.

Evidence:O*NET physicist tasks · NIST AI risk management framework

03

Irreproducible code and simulation

Typical scenario:Software versions, random seeds, parameters, or data cleaning rules are not recorded.

Roles most affected:Computational physicists, data teams

Possible consequence:Others cannot recompute and the results are hard to review.

How to reduce risk:Version-control the code, data dictionary, and configuration, run benchmark tests, and recompute independently.

Evidence:O*NET physicist tasks · NIST AI risk management framework

04

Experiment safety

Typical scenario:Laser, high-voltage, radiation, or cryogenic equipment may carry specific risks depending on the actual project.

Roles most affected:Experimenters, laboratory managers

Possible consequence:Personal injury or equipment damage.

How to reduce risk:Complete institutional training, approval, and protection first; the exercises on this page do not use real hazardous equipment.

Evidence:O*NET physicist tasks · BLS physicists and astronomers outlook

05

Research integrity and evidence extrapolation

Typical scenario:Selective reporting, fabricated citations, or interpreting correlation as physical causation.

Roles most affected:Authors, mentors, peer reviewers

Possible consequence:Conclusions are unreliable and harm the research and its subsequent application.

How to reduce risk:Publish the methods, negative results, and limitations, verify cited originals, and respond to review.

Evidence:O*NET physicist tasks · NIST AI risk management framework

06

Project and career uncertainty

Typical scenario:Research funding, project deadlines, and position degree requirements vary by institution.

Roles most affected:Students, job seekers, research teams

Possible consequence:Investment and position demand may be mismatched.

How to reduce risk:Verify specific hiring, funding deadlines, and transferable skills; do not treat the BLS projection as a personal employment promise.

Evidence:O*NET physicist tasks · BLS physicists and astronomers outlook

Entering the field, qualifications, and career development

How do I enter the physicist occupation?

Build a foundation in mathematics, physics, programming, and experiments during the bachelor's degree, and learn research questions, reproducible code, and error analysis through mentor projects. BLS states that research and university physics positions usually require a doctorate and some federal entry positions may accept a bachelor's degree; the specific position's degree, research direction, and equipment safety requirements depend on the posting.

Start by verifying courses, projects, and position requirements

How do the entry positions and initial tasks of physicists compare?

Entry positionsWhat you mainly do at the startWhat employers usually look atHow to prove it without experienceDifferences from other directions
Undergraduate physics project assistantsClean public or authorized data, recompute basic models, and record experimental conditions under mentor supervision.Mathematics, physics, programming, and record-keeping habits; look at the specific position.Submit a small reproducible project with units, methods, and errors.Assist with research without independently defining project conclusions.
Federal entry-level physics positionsAssist with measurement, calculation, or research material organization as the position requires.BLS states that some federal entry positions may accept a bachelor's degree; the specific position still requires verification.Use courses and projects to demonstrate theoretical, experimental, or computational skills.Different from the doctorate requirement for independent university research positions.
Graduate research assistantsDesign part of the methods, analyze data, and write reports in a mentor's project.Research direction, prerequisite courses, and project fit.Show models, code, errors, and feedback revisions.Under mentor supervision, with duties and funding depending on the project.
Postdoctoral or junior researchersAdvance an independent problem and publish collaboratively with the team.Doctoral training, research outputs, and technical expertise.Papers, reproducible materials, and collaboration records.Usually oriented toward a research career, with contract and funding terms requiring separate verification.
Corporate applied physics R&DAnalyze material or device performance and take part in simulation and experimental validation.The physics direction, engineering constraints, and project experience required by the position.Show performance metrics, boundary conditions, and validation reports.Final deliverables are closer to product technical goals.

What do employers usually look at in hiring?

Degree in the posting

BLS states that research and university positions usually require a doctorate; some federal entry positions may accept a bachelor's degree.

Verify the target posting's degree, direction, and duties item by item.

Course and project ability

Use reproducible analysis of public data to demonstrate dimensions, modeling, programming, and error handling.

State which parts you completed yourself and which were mentor-guided.

O*NET occupational tasks

Verify whether the target position includes complex computation, experimental equipment, theory, simulation, papers, or teaching.

Every task cannot be inferred to appear just from the title 'physicist'.

Safety and team collaboration

Show data records, versions, peer review, and equipment safety awareness.

Operation of hazardous equipment follows institutional training and authorization.

Source: U.S. Bureau of Labor Statistics|United States|2025 wages; 2025—2035 projection|Official occupational or statistical materials · Source: FermatMind editorial compilation|Editorial synthesis|Read and verified 2026-09-26; see the link for the original year|Editorial synthesis

Entry exercise: testing a physics model with fictional pendulum data

Materials and setup|all fictional pendulum data

You are studying the relationship between pendulum length L and period T. The three sets of numbers below are purely teaching setups, not experimental measurements; use only paper or script calculations and do not build an apparatus.

A1: the reference gravitational acceleration is taken as 9.8 meters per second squared; the small-angle ideal model is T = 2 pi times the square root of L over g, ignoring air resistance and amplitude differences.

A2: the fictional data are L = 0.25, 0.36, and 0.49 meters, with measured T = 1.01, 1.20, and 1.42 seconds; the reading uncertainty for each T is provisionally set at plus or minus 0.03 seconds.

Tasks and deliverables

Calculate the three ideal periods, list and compare measured and model values, note meters, seconds, and residuals; plot T squared against L.

State which settings were not given: amplitude, number of timing runs, apparatus calibration; the residuals must not be interpreted as a proven new physical effect.

Reference approach (complete it yourself first)

The model periods are about 1.00, 1.20, and 1.40 seconds; the third residual is about +0.02 seconds, falling within the given plus or minus 0.03 second reading uncertainty.

T squared is approximately linear in L, but with only three fictional points the model cannot be proven to hold at any pendulum length and amplitude; before real measurement, a calibration and repetition plan should be proposed.

Boundaries and review

Review units, formulas, squaring, and rounding; keep the script and the original settings. The exercise requires no laser, high voltage, or any real experimental apparatus.

Source: FermatMind editorial compilation|Editorial synthesis|Read and verified 2026-09-26; see the link for the original year|Editorial synthesis

What reviewable work materials can be prepared at entry?

Question and hypothesis package

Record the ideal pendulum model's variables, units, boundaries, and testable predictions.

Review: others can recompute the three theoretical periods according to A1.

Data and code package

Use A2 to compute T squared, the theoretical T, and the residuals in a list or script, keeping the original fictional values.

Review: units, formulas, and rounding rules are transparent.

Error and limitation package

Explain the plus or minus 0.03 second reading uncertainty and the missing amplitude and calibration information.

Result explanation card

State clearly which observations support the model, which cannot prove it, and propose the next round of measurement planning.

Source: FermatMind editorial compilation|Editorial synthesis|Read and verified 2026-09-26; see the link for the original year|Editorial synthesis

What might the interview and probation period examine, and which conclusions cannot be inferred?

StagePossible focusObservation evidenceUse boundaries
Interview follow-up on the modelHow to obtain the relationship between T squared and L from T = 2 pi times the square root of L over gAble to derive T squared = 4 pi squared L over g and explain the small-angle conditionA question from a single employer does not represent an industry examination
Interview follow-up on errorHow to interpret the third residualFirst compare with the plus or minus 0.03 seconds, then check systematic error and sample sizeFictional data cannot be treated as a real experiment
Project handoverWhether others can recompute the resultsProvide units, data, code versions, and calculation stepsReproducibility must be verified separately on real data
Experiment safetyWhether one can directly handle unknown equipmentFirst verify training, authorization, and risk controlPaper exercises do not grant equipment operating rights
Source: FermatMind editorial compilation|Editorial synthesis|Read and verified 2026-09-26; see the link for the original year|Editorial synthesis · Source: U.S. Bureau of Labor Statistics|United States|2025 wages; 2025—2035 projection|Official occupational or statistical materials

Seven-day experience: recomputing models, errors, and research conclusions

  1. 01

    Day One

    Read O*NET 19-2012 and the BLS education note and write the differences among theoretical, experimental, and computational tasks.

  2. 02

    Day Two

    Copy the A1 model, verify the units of L, g, and T, and list the small-angle and idealization assumptions.

  3. 03

    Day Three

    Compute the three theoretical periods according to A2; use paper or code and keep the calculation steps.

  4. 04

    Day Four

    Compute T squared, the residuals, and the error interval, and mark which differences cannot support a conclusion.

  5. 05

    Day Five

    Plot T squared against L and write a paragraph on the range where the model and data agree and the missing measurement conditions.

  6. 06

    Day Six

    Seal the first draft first, then compare with the reference approach; record the reasons for revisions to formulas, units, or rounding.

  7. 07

    Day Seven

    Prepare a one-page reproducible report, ask a physics teacher for feedback, and record the effort and difficulties without giving a fit score.

Source: FermatMind editorial compilation|Editorial synthesis|Read and verified 2026-09-26; see the link for the original year|Editorial synthesis

How to decide at the end of seven days whether to continue

  • Keep the first draft and the revised draft and record the model assumptions, calculation errors, and revision basis item by item.
  • Check whether fictional data were mistakenly called measured data, or the plus or minus 0.03 second reading uncertainty was treated as the whole error.
  • Record the daily effort and energy feelings and describe only the changes, without converting them into a fit score.
  • Recompute the three theoretical periods and check whether the units, squaring, and rounding are correct.
  • If it is still unclear whether you prefer theory or experiment, list specific questions and ask a teacher or researcher.
  • Compare the report with real position requirements and identify the courses, projects, or degrees still missing.
  • Record unfinished parts honestly and do not fill gaps with conclusions.
Source: FermatMind editorial compilation|Editorial synthesis|Read and verified 2026-09-26; see the link for the original year|Editorial synthesis

Education, training, and qualifications: which are the real thresholds for this occupation?

Qualification / identityWho it mainly suitsRealistic role in job seekingWhether it is a general hiring thresholdApplication / registration conditionsLevel of investmentImportant limits
Undergraduate physics degreePeople who want to build a physics and mathematics foundationSupports some entry position applications and graduate prerequisitesNot a sufficient condition for all research positionsVerify against the school curriculum and the target positionYears of course and project investmentA degree does not automatically grant experimental equipment access
Research master'sPeople who want to deepen a direction or move into applied researchMay support some R&D positions; see the postingNo unified occupational thresholdVerify the mentor, curriculum, thesis, and project conditionsUsually requires a long time and fundingNot equal to the doctorate required for independent research positions
Physics doctorateApplicants for research or university positionsBLS states that such positions usually require a doctorateA typical path, but not required for all physics workApply according to program, mentor, and school requirementsLong-term research trainingA doctorate does not guarantee a position or funding
Laboratory safety trainingResearchers who actually operate equipmentMake operations meet institutional rules and project risk requirementsDecided by the specific equipment and institutionFirst verify the equipment, training, and authorizationContinuous retraining depending on the projectA course certificate is not universal authorization for all equipment
Programming and statistics projectsPeople in computational or experimental data directionsShow code, error analysis, and reproducible resultsNot a unified practice qualificationVerify data permissions and the project sourceCan be built through courses and public projectsWork must state original contributions and limitations
Papers and peer reviewResearch position applicantsShow methods, conclusions, and the review processRequirements differ by positionFollow research ethics and submission rulesYears of accumulationThe number of papers is not equal to the quality of conclusions

Notes on qualifications and authorization boundaries

Notes on the hiring sample

Theoretical and computational physics

01

Undergraduate exploration

Typical scope:Learn mathematics and physics foundations and experience mathematical models, derivations, and simulations.

Readiness evidence:Courses and a small public-data project.

Credential boundary:Undergraduate course requirements depend on the school; there is no unified practice certificate.

Next moves:Choose courses and mentor projects and verify the target position.

02

Project participants

Typical scope:Take on a defined part of mathematical models, derivations, and simulations under mentor supervision.

Readiness evidence:Project notes, consistency between predictions and observations, and mentor feedback.

Credential boundary:Train according to project and equipment safety requirements; do not independently undertake hazardous equipment operations.

Next moves:Decide whether to apply for a research graduate program.

03

Graduate students or junior researchers

Typical scope:Design and review mathematical models, derivations, and simulations and explain uncertainty.

Readiness evidence:Reproducible code, consistency between predictions and observations, and paper drafts.

Credential boundary:Research and university positions usually require a doctorate; specific positions require separate verification.

Next moves:Verify the doctoral research direction, position, and funding conditions.

04

Independent researchers

Typical scope:Pose questions, coordinate mathematical models, derivations, and simulations, and take responsibility for the results.

Readiness evidence:Papers, peer review, and ongoing records of consistency between predictions and observations.

Credential boundary:According to the institutional position, project, and equipment authorization; a degree does not automatically authorize all operations.

Next moves:Continuously review limiting cases, numerical convergence, and consistency with observations; interdisciplinary or management duties require separate verification.

Experimental physics

01

Undergraduate exploration

Typical scope:Learn mathematics and physics foundations and experience instruments, measurement, calibration, and errors.

Readiness evidence:Courses and a small public-data project.

Credential boundary:Undergraduate course requirements depend on the school; there is no unified practice certificate.

Next moves:Choose courses and mentor projects and verify the target position.

02

Project participants

Typical scope:Take on a defined part of instruments, measurement, calibration, and errors under mentor supervision.

Readiness evidence:Project notes, reproducible raw data, and mentor feedback.

Credential boundary:Train according to project and equipment safety requirements; do not independently undertake hazardous equipment operations.

Next moves:Decide whether to apply for a research graduate program.

03

Graduate students or junior researchers

Typical scope:Design and review instruments, measurement, calibration, and errors and explain uncertainty.

Readiness evidence:Reproducible code, reproducible raw data, and paper drafts.

Credential boundary:Research and university positions usually require a doctorate; specific positions require separate verification.

Next moves:Verify the doctoral research direction, position, and funding conditions.

04

Independent researchers

Typical scope:Pose questions, coordinate instruments, measurement, calibration, and errors, and take responsibility for the results.

Readiness evidence:Papers, peer review, and ongoing records of reproducible raw data.

Credential boundary:According to the institutional position, project, and equipment authorization; a degree does not automatically authorize all operations.

Next moves:Continuously review calibration drift, raw records, and the error budget, and train the team in safety handover; interdisciplinary or management duties require separate verification.

Applied physics and R&D

01

Undergraduate exploration

Typical scope:Learn mathematics and physics foundations and experience applying physics methods to devices or technical solutions.

Readiness evidence:Courses and a small public-data project.

Credential boundary:Undergraduate course requirements depend on the school; there is no unified practice certificate.

Next moves:Choose courses and mentor projects and verify the target position.

02

Project participants

Typical scope:Take on a defined part of applying physics methods to devices or technical solutions under mentor supervision.

Readiness evidence:Project notes, performance, constraints, validation results, and mentor feedback.

Credential boundary:Train according to project and equipment safety requirements; do not independently undertake hazardous equipment operations.

Next moves:Decide whether to apply for a research graduate program.

03

Graduate students or junior researchers

Typical scope:Design and review the application of physics methods to devices or technical solutions and explain uncertainty.

Readiness evidence:Reproducible code, performance, constraints, validation results, and paper drafts.

Credential boundary:Research and university positions usually require a doctorate; specific positions require separate verification.

Next moves:Verify the doctoral research direction, position, and funding conditions.

04

Independent researchers

Typical scope:Pose questions, coordinate the application of physics methods to devices or technical solutions, and take responsibility for the results.

Readiness evidence:Papers, peer review, and ongoing records of performance, constraints, and validation results.

Credential boundary:According to the institutional position, project, and equipment authorization; a degree does not automatically authorize all operations.

Next moves:Review the technical solution against performance metrics, physical boundaries, and validation results; interdisciplinary or management duties require separate verification.

Capability ladder: what demonstrates readiness?

  1. 01

    Exploration

    Verify the mathematics, physics, and programming course requirements and recompute simple models with public data.

  2. 02

    Foundation training

    Take core physics courses and learn experimental records, statistics, code, and error analysis.

  3. 03

    Research training

    Join a mentor's project, define the question, methods, and original contribution, and accept peer feedback.

  4. 04

    Independent responsibility

    Design research, review evidence, publish or apply results, and train the team within the position scope.

Career prospects and related occupation transitions

How should physicist market signals be read?

BLS Table 1.2 United States 19-2012 physicists 23,200 in 2025 and 24,800 in 2035, overall growth 7.2%; the original table's ten-year net increase is 1,700, and the displayed endpoint difference of 1,600 cannot replace the unrounded net increase. Average annual openings 1,300 including growth and replacements, not a net increase or immediate hiring; the overall projection is not attributed to AI and does not guarantee opportunities in China or a specific research direction.

Three sources, three different measures

BLS detailed occupational projections

2025

United States · 19-2012 physicists

About 23,200 people

Employment base

The United States employment base period of the BLS 19-2012 standalone row.

Does not establish:Not equal to the current number of hires.

BLS detailed occupational projections

2025—2035

United States · 19-2012 physicists

7.2%

Projected growth

BLS Table 1.2 United States 19-2012 overall growth 7.2%, from 23,200 to 24,800; the original table's ten-year net increase is 1,700, not attributed to AI.

Does not establish:The overall non-AI projection cannot infer Chinese demand, research direction, or individual hiring.

BLS detailed occupational projections

2025—2035

United States · 19-2012 physicists

About 1,300 per year

Average annual openings

Includes growth and replacement demand from role changes, retirement, and the like.

Does not establish:1,300 per year includes growth and replacements, not a ten-year net increase of 1,700 or immediate hiring, and not attributed to AI.

Where can your current capabilities transfer?

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

Astronomers

Editorial comparison: medium

Transfers:Mathematical modeling and observational data analysis

Build next:Astrophysics domain knowledge, observational equipment, and research topics

View career profile

Materials scientists

Editorial comparison: medium

Transfers:Physical measurement, statistics, and material characterization

Build next:Material synthesis, performance, and processing knowledge

View career profile

Mathematicians

Editorial comparison: medium

Transfers:Mathematical modeling and computation

Build next:Rigorous proofs and mathematical theory outputs

View career profile

Nuclear engineers

Editorial comparison: high

Transfers:Nuclear physics foundations, modeling, and measurement

Build next:Engineering design, safety codes, and institutional authorization

View career profile

Computer and information research scientists

Editorial comparison: medium

Transfers:Algorithms, numerical methods, and programming

Build next:Computing system or algorithm research questions

View career profile

Postsecondary physics teachers

Editorial comparison: medium

Transfers:Physics knowledge, research, and expression

Build next:Instructional design, the doctorate, and university position requirements

View career profile

Electrical engineers

Editorial comparison: high

Transfers:Electromagnetism, instruments, and simulation

Build next:Electrical system design, codes, and product delivery

View career profile

Data scientists

Editorial comparison: medium

Transfers:Statistics, programming, and data analysis

Build next:Business problems, data products, and deployment evaluation

View career profile

Frequently asked questions and sources

What do physicists do every day?

Study physics problems, build mathematical models, design experiments or simulations, analyze data and errors, review conclusions, and write papers or technical reports. O*NET 19-2012 lists these tasks; theory, experiment, and application positions combine them differently.

What is the difference between physicists and engineers?

Physicists usually take explaining and predicting physical phenomena as their main output, while engineers focus more on delivering systems under technical specifications, cost, and safety constraints; actual positions overlap, so the job duties must be checked.

How do I judge whether I am suited?

See whether you are willing to work long term on mathematical reasoning, experimental or computational verification, error analysis, and repeated recomputation. The O*NET interest code IRC is for exploration only, not an ability or hiring threshold.

Will AI replace physicists?

No 19-2012-specific AI position replacement rate was verified this time. AI can assist with literature, code, and data organization, and conclusions must still be checked for dimensions, raw data, model boundaries, and reproducibility.

What is the national wage median for Chinese physicists?

No national Chinese physicist wage median directly comparable to United States 19-2012 was verified. According to the research data, computational models, experiment, and paper tasks, refer to the 2025 national scientific research and technical services pre-tax annual averages: non-private 182,064 and private 83,560 yuan per year; 155,491 yuan is the cross-industry broad professional technical position average for above-scale enterprises. None is the physics occupational median or the complete university, institute, or postdoctoral market, and specific positions and contracts require separate verification.

Is the United States median wage new entrant income?

No. The United States 19-2012 median annual wage in 2025 was 172,250 dollars, and the occupational percentiles are not grouped by new entrants, region, or research direction.

Must physicists have a doctorate?

BLS states that research and university physics positions usually require a doctorate and some federal entry positions may accept a bachelor's degree; specific employer and research direction requirements require separate verification.

How do I choose between theoretical and experimental physics?

Theoretical and computational directions emphasize models, derivation, and simulation, while experimental directions emphasize instruments, calibration, and measurement; both need testable results and uncertainty to constrain conclusions.

What safe exercises can I do at entry?

Use public or fictional pendulum data to compute model periods and residuals, state the units, assumptions, and errors clearly, and ask a teacher or researcher for feedback; do not operate hazardous equipment.

What are the employment prospects for United States physicists?

BLS Table 1.2 United States 19-2012 physicists 23,200 in 2025 and 24,800 in 2035, overall growth 7.2%; the original table's ten-year net increase is 1,700, and the displayed endpoint difference of 1,600 cannot replace the unrounded net increase. Average annual openings 1,300 including growth and replacements, not a net increase or immediate hiring; the overall projection is not attributed to AI and does not guarantee opportunities in China or a specific research direction.

Sources

FermatMind career content team (editorial compilation, not reviewed by practicing physicists)

Material review date on this page: 2026-09-26. Occupational tasks, wages, employment, and education paths should be re-checked against the listed official sources and the actual job posting.

O*NET19-2012 supports physics tasks and IRC; the BLS standalone row supports 2025 wages, the four industry medians, and overall employment, and the physics and astronomy group supports education. The NBS scientific research industry and broad professional technical positions are only Chinese wage background, not a physics occupational median.

  • O*NET Physicists 19-2012.00U.S. Department of Labor, O*NET|United States|Read and verified 2026-09-26; see the link for the original year|Official occupational or statistical materials
    View original source:O*NET Physicists 19-2012.00
    • Physicist definition, data analysis, complex computation, experiment, theory, and research communication
    • The tasks come from an occupational survey and do not mean every research direction has the same frequency

    Limitations:The tasks come from an occupational survey and do not mean every research direction has the same frequency

  • contextU.S. Bureau of Labor Statistics|United States|2025 wages; 2025—2035 projection|Official occupational or statistical materials
    View original source:context
    • Physics and astronomy education background: research and university positions usually require a doctorate, and federal entry physics positions usually require a bachelor's degree.
    • 2025 physicist industry medians: state universities 84600, scientific research and development 176170, federal government 146900, and outpatient medical care 239360 dollars per year; not astronomer wages.

    Limitations:The group narrative does not mean every physicist needs a doctorate

  • BLS 19-2012 occupational projectionsU.S. Bureau of Labor Statistics|United States|2025 base period; 2025—2035|Official occupational or statistical materials
    View original source:BLS 19-2012 occupational projections
    • The 19-2012 standalone row 23.2 thousand to 24.8 thousand, overall 7.2%, an original-table ten-year net increase of 1.7 thousand, and 1.3 thousand annually including replacements.
    • The displayed endpoint difference of 1.6 thousand does not replace the original table's 1.7 thousand, is not attributed to AI, and does not guarantee China or a specific research direction.

    Limitations:Annual openings include replacements and differ from the ten-year net increase of about 1,700

  • O*NET/BLS 2025 Physicists wagesU.S. Department of Labor, O*NET|United States|May 2025 wages|Official occupational or statistical materials
    View original source:O*NET/BLS 2025 Physicists wages
    • The national 19-2012 annual wage 10 / 25 / 50 / 75 / 90 percentiles are 82110 / 111720 / 172250 / 222760 / 274110 dollars.
    • The median hourly wage is 82.81 dollars; the percentiles are not a new entrant starting salary, individual after-tax pay, or a Chinese quote.

    Limitations:Percentiles cannot be treated as a new graduate starting salary or a Chinese wage

  • contextNational Bureau of Statistics|Mainland China|Annual 2025; released May 2026|Official occupational or statistical materials
    View original source:context
    • NBS tables 2 and 5 scientific research and technical services non-private 182064 and private 83560 yuan per year, pre-tax industry averages for all positions.
    • 155491 is the cross-industry broad professional technical position average for above-scale enterprises, not the physicist median or the complete university, institute, and postdoctoral market.

    Limitations:Urban legal entities with 5 or more employees, excluding the self-employed and freelancers; the industry's individual positions and the broad professional technical position group differ, and they cannot be pieced into an occupational wage range.

  • contextU.S. Department of Labor, O*NET|Mainland China|Read and verified 2026-09-26; see the link for the original year|Editorial synthesis
    View original source:context
    • United States physicist tasks and classification boundaries; cross-referenced to Chinese positions for comparison only
    • Does not provide a unified qualification or salary for Chinese research positions

    Limitations:Does not provide a unified qualification or salary for Chinese research positions

  • contextNational Institute of Standards and Technology|Cross-region|Read and verified 2026-09-26; see the link for the original year|Research or governance materials
    View original source:context
    • Organization-level generative AI risk management
    • Does not validate the correctness of a specific scientific model

    Limitations:Does not validate the correctness of a specific scientific model

  • O*NET Physicists 19-2012.00U.S. Department of Labor, O*NET|United States|Read and verified 2026-09-26; see the link for the original year|Official occupational or statistical materials
    View original source:O*NET Physicists 19-2012.00
    • Physicist definition, data analysis, complex computation, experiment, theory, and research communication
    • The tasks come from an occupational survey and do not mean every research direction has the same frequency

    Limitations:The tasks come from an occupational survey and do not mean every research direction has the same frequency

  • O*NET Physicists 19-2012.00U.S. Department of Labor, O*NET|United States|Read and verified 2026-09-26; see the link for the original year|Official occupational or statistical materials
    View original source:O*NET Physicists 19-2012.00
    • Physicist definition, data analysis, complex computation, experiment, theory, and research communication
    • The tasks come from an occupational survey and do not mean every research direction has the same frequency

    Limitations:The tasks come from an occupational survey and do not mean every research direction has the same frequency

  • O*NET Physicists 19-2012.00U.S. Department of Labor, O*NET|United States|Read and verified 2026-09-26; see the link for the original year|Official occupational or statistical materials
    View original source:O*NET Physicists 19-2012.00
    • Physicist definition, data analysis, complex computation, experiment, theory, and research communication
    • The tasks come from an occupational survey and do not mean every research direction has the same frequency

    Limitations:The tasks come from an occupational survey and do not mean every research direction has the same frequency

  • contextU.S. Department of Labor, O*NET|Mainland China|Read and verified 2026-09-26; see the link for the original year|Editorial synthesis
    View original source:context
    • United States physicist tasks and classification boundaries; cross-referenced to Chinese positions for comparison only
    • Does not provide a unified qualification or salary for Chinese research positions

    Limitations:Does not provide a unified qualification or salary for Chinese research positions

  • contextInternational Labour Organization|Cross-region|Read and verified 2026-09-26; see the link for the original year|Research or governance materials
    View original source:context
    • Cross-occupation generative AI task exposure
    • Does not measure physics research quality or net positions
    • Do not infer Chinese income or AI replacement rates from this source.

    Limitations:Does not measure physics research quality or net positions

  • editorialFermatMind editorial compilation|Editorial synthesis|Read and verified 2026-09-26; see the link for the original year|Editorial synthesis
    • Fictional physics experiment data exercises and source interpretation
    • Not a formal research project or laboratory qualification

    Limitations:Not a formal research project or laboratory qualification

  • contextNational Institute of Standards and Technology|Cross-region|Read and verified 2026-09-26; see the link for the original year|Research or governance materials
    View original source:context
    • Organization-level generative AI risk management
    • Does not validate the correctness of a specific scientific model

    Limitations:Does not validate the correctness of a specific scientific model

  • contextInternational Labour Organization|Cross-region|Read and verified 2026-09-26; see the link for the original year|Research or governance materials
    View original source:context
    • Cross-occupation generative AI task exposure
    • Does not measure physics research quality or net positions

    Limitations:Does not measure physics research quality or net positions

  • contextInternational Labour Organization|Cross-region|Read and verified 2026-09-26; see the link for the original year|Research or governance materials
    View original source:context
    • Cross-occupation generative AI task exposure
    • Does not measure physics research quality or net positions

    Limitations:Does not measure physics research quality or net positions

  • contextU.S. Bureau of Labor Statistics|United States|2025 wages; 2025—2035 projection|Official occupational or statistical materials
    View original source:context
    • The physicist work environment and the research, academic, and federal position education paths
    • The group narrative does not mean every physicist needs a doctorate

    Limitations:The group narrative does not mean every physicist needs a doctorate

  • O*NET Physicists 19-2012.00U.S. Department of Labor, O*NET|United States|Read and verified 2026-09-26; see the link for the original year|Official occupational or statistical materials
    View original source:O*NET Physicists 19-2012.00
    • Physicist definition, data analysis, complex computation, experiment, theory, and research communication
    • The tasks come from an occupational survey and do not mean every research direction has the same frequency

    Limitations:The tasks come from an occupational survey and do not mean every research direction has the same frequency

  • O*NET/BLS 2025 Physicists wagesU.S. Department of Labor, O*NET|United States|May 2025 wages|Official occupational or statistical materials
    View original source:O*NET/BLS 2025 Physicists wages
    • The national 19-2012 annual wage 10 / 25 / 50 / 75 / 90 percentiles are 82110 / 111720 / 172250 / 222760 / 274110 dollars.
    • The median hourly wage is 82.81 dollars; the percentiles are not a new entrant starting salary, individual after-tax pay, or a Chinese quote.

    Limitations:Percentiles cannot be treated as a new graduate starting salary or a Chinese wage

  • O*NET/BLS 2025 Physicists wagesU.S. Department of Labor, O*NET|United States|May 2025 wages|Official occupational or statistical materials
    View original source:O*NET/BLS 2025 Physicists wages
    • The national 19-2012 annual wage 10 / 25 / 50 / 75 / 90 percentiles are 82110 / 111720 / 172250 / 222760 / 274110 dollars.
    • The median hourly wage is 82.81 dollars; the percentiles are not a new entrant starting salary, individual after-tax pay, or a Chinese quote.

    Limitations:Percentiles cannot be treated as a new graduate starting salary or a Chinese wage

  • O*NET/BLS 2025 Physicists wagesU.S. Department of Labor, O*NET|United States|May 2025 wages|Official occupational or statistical materials
    View original source:O*NET/BLS 2025 Physicists wages
    • The national 19-2012 annual wage 10 / 25 / 50 / 75 / 90 percentiles are 82110 / 111720 / 172250 / 222760 / 274110 dollars.
    • The median hourly wage is 82.81 dollars; the percentiles are not a new entrant starting salary, individual after-tax pay, or a Chinese quote.

    Limitations:Percentiles cannot be treated as a new graduate starting salary or a Chinese wage

  • contextNational Bureau of Statistics|Mainland China|Annual 2025; released May 2026|Official occupational or statistical materials
    View original source:context
    • NBS tables 2 and 5 scientific research and technical services non-private 182064 and private 83560 yuan per year, pre-tax industry averages for all positions.
    • 155491 is the cross-industry broad professional technical position average for above-scale enterprises, not the physicist median or the complete university, institute, and postdoctoral market.

    Limitations:Urban legal entities with 5 or more employees, excluding the self-employed and freelancers; the industry's individual positions and the broad professional technical position group differ, and they cannot be pieced into an occupational wage range.

  • O*NET Physicists 19-2012.00U.S. Department of Labor, O*NET|United States|Read and verified 2026-09-26; see the link for the original year|Official occupational or statistical materials
    View original source:O*NET Physicists 19-2012.00
    • Physicist definition, data analysis, complex computation, experiment, theory, and research communication

    Limitations:The tasks come from an occupational survey and do not mean every research direction has the same frequency

  • O*NET/BLS 2025 Physicists wagesU.S. Department of Labor, O*NET|United States|May 2025 wages|Official occupational or statistical materials
    View original source:O*NET/BLS 2025 Physicists wages
    • The national 19-2012 annual wage 10 / 25 / 50 / 75 / 90 percentiles are 82110 / 111720 / 172250 / 222760 / 274110 dollars.
    • The median hourly wage is 82.81 dollars; the percentiles are not a new entrant starting salary, individual after-tax pay, or a Chinese quote.

    Limitations:Percentiles cannot be treated as a new graduate starting salary or a Chinese wage

  • O*NET Physicists 19-2012.00U.S. Department of Labor, O*NET|United States|Read and verified 2026-09-26; see the link for the original year|Official occupational or statistical materials
    View original source:O*NET Physicists 19-2012.00
    • Physicist definition, data analysis, complex computation, experiment, theory, and research communication
    • The tasks come from an occupational survey and do not mean every research direction has the same frequency

    Limitations:The tasks come from an occupational survey and do not mean every research direction has the same frequency

  • 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+25+75)/3+0.5)=50; 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.
    • Method design|Choose theoretical derivation, experiment, or simulation|Specify variables, boundary conditions, and controls. — 50/100; weight=1/3; physicists.profile.responsibilities-block.1[1]
    • Data acquisition|Operate equipment or obtain authorized data according to training and rules|Keep calibration, parameters, and raw records. — 25/100; weight=1/3; physicists.profile.responsibilities-block.1[2]
    • Analysis and interpretation|Check dimensions, errors, and alternative explanations|Do not equate a model fit with proof of causation. — 75/100; weight=1/3; physicists.profile.responsibilities-block.1[3]

    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.

  • Occupational tasks; for combined pages applies to the named component occupationO*NET OnLine (U.S. DOL/ETA)|United States|Occupational profile
    View original source:Occupational tasks; for combined pages applies to the named component occupation
    • The task table describes the work and its typical outputs.

    Limitations:Check duties and authorization for the actual workplace.

  • Fictional occupational recordFermatMind|This page|2026-11-10
    • Use a fictional record to identify discrepancies and hand them off for review.

    Limitations:Paper exercise only.

  • 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.

The Chinese text on this page explains the scope of the United States official occupation and does not establish mutual recognition of Chinese and foreign qualifications; all fictional exercise materials are editorial designs and do not represent employer standards or legal requirements. Individual results, orders, and payment information do not enter public pages.

Physicists | FermatMind