EXPOSED
Much of the day — literature review, composition screening, DFT/FEA setup, statistical analysis of characterization data, technical report and patent drafting — is exactly what generative models and ML property-prediction systems (GNoME, MatterGen-style tools) now do at usable quality, and screening thousands of candidate compositions is where AI is strongest. What survives is bench-side: preparing and mounting samples, running and interpreting SEM/TEM/XRD/DSC output on messy real specimens, diagnosing why a production batch failed, and owning the recommendation that a material is fit for a real part. No license gates the work and buyers rarely pay for the individual relationship, so the moat is embodiment plus accountability for expensive material decisions.
Headcount grew steadily across the period.
Median pay $96,810 → $117,790 -2.7% in real terms
This line is counted by the Bureau of Labor Statistics — the one figure on this page that isn't a judgement of ours. Headcount moves on demand, offshoring, demographics and the business cycle, and automation is one term among several, often not the loudest.
So a falling line is not evidence that AI did it, and a rising one is not evidence that it won't. Both happen in this register: some occupations resist automation and shrink anyway, others are highly automatable and keep growing. The marked year is 2020.
BLS projection, 2024–2034
+4.9% 8,700 → 9,100 on the projections basis
Exposed, but growing
AI can already do a lot of these tasks, and the BLS still expects +4.9% more of these jobs by 2034. Demand for the output is growing faster than the work is being automated away — the mechanism BLS gives for software developers, and the combination people most often misread as an error.
Different clocks. The score is what current AI could do to this work today. The projection is how many of these jobs will exist in 2034. Everything between the two — how fast employers actually adopt, whether demand grows in the meantime — is why they can point opposite ways without either being wrong.
~600 openings a year on average, including replacing people who leave.
ScientistMetallurgistPlastics ScientistPolymer SpecialistResearch ScientistMaterials ScientistAnalytical ScientistMetal Alloy ScientistApplications ScientistMetallurgical EngineerMaterial Science EngineerMaterials Research EngineerPolymer Materials ConsultantResearch Development SpecialistR and D Scientist (Research and Development Scientist)Micro Electrical/Mechanical Systems Device Scientist (MEMS Device Scientist)
Holding it up: judgment & accountability . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier At 10, the split is real: composition screening, phase-diagram lookup, Arrhenius fits, and DSC/TGA curve deconvolution are already handed to ML surrogates and autoscript pipelines, while polishing and ion-milling a TEM lamella, deciding a XRD peak is texture rather than a second phase, and tracing a delamination back to a cure-cycle drift still need someone at the instrument — neither half dominates enough to push this to 7 or to 14.
Some physical or field component 11 reflects that the fume hood, glovebox, furnace, and load frame are unavoidable — you mix slurries, sinter pellets, mount and etch cross-sections, and pull tensiles — but it is your own lab or a plant QA bay with known hazards and fixturing, not a refinery scaffold or a field failure site, which is what keeps it out of the 13+ band.
No licence, no signature requirement 3 is right because nothing stops a chemistry PhD from specifying an alloy: there is no state licensure for materials scientists, and where a stamp is legally required — pressure vessels, structural members — a licensed PE or the ASME/AWS-certified inspector signs, not you, so ASM or NACE credentials read as resume weight rather than a legal gate.
Meaningful discretion 12 sits where it does because you decide whether a lot with borderline grain size ships, whether a substitution qualifies without full requalification, and whether a fractograph is fatigue or overload — calls with real cost and safety consequence — yet they are made inside ASTM/ISO test methods and a customer spec, and the release decision usually clears a design authority or quality board rather than resting on you alone.
The verdict above describes this occupation as a whole. Almost nobody does the typical version of a job — tick what's actually in your week and see how your own mix sits.
Your task mix speaks to task resistance (10/20 here) — how much of the day's work current AI already does. That is the dimension the boxes above are about.
It cannot move the other three. Liability shield (3/20) is whether the law requires a licensed human to sign. Trust premium (7/20) is whether buyers specifically pay for a person. Judgment and accountability (12/20) is whether the role exists to own consequential calls. Those are facts about the occupation's standing, not about which tasks are in your week — a paralegal who does only trial exhibits still holds no licence. Together they are 22 of this occupation's 43 points (51%).
Embodiment (11/20) is also a property of the work rather than the worker, but we don't tag individual tasks as physical or not, so the picker can't tell you anything about it. That's a limit of this tool, not a claim.
Did we get the list right? Tell us what's missing — the tasks are written from the outside, and you're reading this from the inside.
Nuclear Engineers EXPOSED
The moves above are yours to make. This is the other half: what would have to change in the world for the occupation itself to score higher. None of it is in any one person's gift, but it is where the floor actually comes from. Scores here are not a one-way ratchet. Only two of the five dimensions — task resistance and embodiment — track what machines can do. The other three track law, what buyers will pay for, and who is answerable, and those move in both directions, often in response to the same pressure AI creates. If every lever below landed, this occupation would score around 61/100, still EXPOSED.
As AI absorbs composition screening, literature review and simulation setup, the residual job concentrates on failure analysis of real specimens — fractography, contamination diagnosis on production batches, and reconciling characterization artifacts (XRD peak overlap, TEM sample-prep damage) against model predictions. This tier shift is visible already in aerospace/semiconductor labs where autonomous screening platforms feed human root-cause teams, and raises task_resistance without any new rule.
If qualification regimes keep requiring physical coupon testing on the actual production lot — e.g. AMS/NADCAP audit rules and FAA/EASA material qualification requiring witnessed specimen preparation and destructive test, plus semiconductor fab requirements for on-tool cross-section work — the hands-on specimen and metrology fraction of the day stays load-bearing even as design work automates.
ASME BPVC Section IX and AWS weld/material qualifications already require a named responsible engineer to certify procedure qualification records; extension of PE-stamp requirements to material substitution and additive-manufacturing part qualification (NIST/ASTM F42 and ongoing state board debate over AM part certification) would put a personally liable signature on materials fitness calls. Also watch EU Construction Products Regulation and battery passport rules naming a responsible technical person for declared performance.
If insurers and OEMs require a named materials engineer of record for AI-proposed material substitutions — mirroring how aerospace primes already require a named cognizant engineer to disposition nonconformances (MRB authority) — and if AI-generated candidate materials must be dispositioned by that person before release, the role formally owns consequential ambiguous calls.
Narrow route only: expert-witness and independent failure-analysis work (product liability, battery fire and construction-defect litigation), where court admissibility under FRE 702 and Daubert requires a testifying human whose own methodology is examinable. This is a small slice of the occupation and will not lift the median role.
The limit. No license gates the title and most materials scientists are salaried employees inside firms that buy outcomes, not individual relationships — so trust_premium has a hard low ceiling. Liability routes attach to a subset (pressure vessels, aerospace, construction products); a battery or thin-film researcher gains nothing from them.
| Boston-Cambridge-Newton, MA-NH | 700 | $132,600 +13% |
| New York-Newark-Jersey City, NY-NJ | 560 | $120,650 +2% |
| Chicago-Naperville-Elgin, IL-IN | 460 | $133,340 +13% |
| San Francisco-Oakland-Fremont, CA | 380 | $231,710 +97% |
| Minneapolis-St. Paul-Bloomington, MN-WI | 270 | $132,160 +12% |
| Atlanta-Sandy Springs-Roswell, GA | 240 | $93,240 -21% |
| Denver-Aurora-Centennial, CO | 240 | $106,720 -9% |
| Cincinnati, OH-KY-IN | 200 | $107,500 -9% |
| San Jose-Sunnyvale-Santa Clara, CA | 180 | $231,920 +97% |
| San Francisco-Oakland-Fremont, CA | 380 | $231,710 +97% |
| Worcester, MA | 80 | $151,070 +28% |
New York State announced that Radical AI has established what it describes as the state's first fully autonomous materials science laboratories at the Brooklyn Navy Yard, using AI and robotics to run experiments.
Has AI actually changed your work? One tap, anonymous, and the running tally is public. Nothing else is asked of you.
Rather than check back: get the digest and we'll tell you what changed — or watch a single occupation from its own page.