EXPOSED
Materials engineers split their week between screen work that AI is rapidly absorbing — literature review, materials property lookups, spec drafting, simulation setup, test-report writing — and physical work that it can't touch: pulling failed parts, running SEM and tensile tests, standing on a plant floor diagnosing why a heat treat batch went brittle. ML models for alloy and polymer property prediction are already displacing chunks of candidate screening that used to be human days. What holds is failure-analysis judgment on real hardware and ownership of a materials selection call when a part cracks in the field; PE licensure exists but is rarely required in this discipline, so the regulatory moat is thin.
Nearly all of this fall was the 2020 shock. It has been climbing back since.
Median pay $93,360 → $112,860 -3.3% 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
+5.7% 23,000 → 24,300 on the projections basis
Exposed, but growing
AI can already do a lot of these tasks, and the BLS still expects +5.7% 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.
~1,500 openings a year on average, including replacing people who leave.
EngineerMetallurgistTest EngineerMetallographerDesign EngineerStress EngineerCeramic EngineerMaterial AnalystPlating EngineerPolymer EngineerProject EngineerTesting EngineerWelding EngineerPlastics EngineerResearch EngineerSmelting EngineerCorrosion EngineerExtrusion EngineerMaterials EngineerFoundry MetallurgistProcess MetallurgistPhysical MetallurgistCeramics Test EngineerGlass Science Engineer
Holding it up: judgment & accountability . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier Composition screening, phase-diagram lookups, ASTM spec drafting and FEA/CALPHAD setup are already being compressed by property-prediction models trained on the same handbook data you use, but mounting and polishing a fracture surface, interpreting an SEM fractograph against the actual load history, and correlating a supplier's mill cert with a field failure still require someone in the lab — roughly half the week resists, which is why this sits at 11 and not 15.
Some physical or field component You are in a controlled lab most of the time — tensile frames, DSC, metallography, salt-spray cabinets — with periodic trips to a plant floor for heat-treat, weld, or coating-line troubleshooting; that mix of benchtop instruments plus occasional uncontrolled production environments is what puts this at 9 rather than the 3 of a pure simulation role or the 15 of a field inspector.
Certification preferred, not legally required There is no licence gating materials work: PE in metallurgical/materials is offered but almost no employer or contract demands it, and ASM/NACE/AWS certifications (CWI, cathodic protection) are asked for in specific niches rather than across the job, so the 5 reflects credentials that help you get hired, not statutes that make you personally answerable.
Meaningful discretion Choosing 17-4PH over 316L for a corrosive fatigue application, calling whether hydrogen embrittlement or a forging lap caused the crack, and deciding a nonconforming lot ships or scraps are consequential calls made with incomplete data — real ownership, but bounded by ASTM/AMS specs, customer drawings and a design authority above you, which is why it lands at 12 instead of 16.
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 (11/20 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 (5/20) 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 24 of this occupation's 44 points (55%).
Embodiment (9/20) 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 60/100, still EXPOSED.
AS9100/IATF 16949 or Nadcap audit criteria requiring a named human engineer to review and sign any AI-generated property prediction or lifing analysis before it enters a certified design record — analogous to how Nadcap already requires named-approved personnel for heat treat pyrometry.
Task-mix shift: this occupation genuinely has two tiers. If ML surrogate models absorb candidate screening, literature review and simulation setup, the residual day becomes root-cause failure analysis on real fracture surfaces, correlating fractography with process records, and defending conclusions in warranty or recall proceedings — work that is currently hard to do without hardware access and chain-of-custody discipline.
Aerospace/nuclear qualification regimes tightening the named-engineer requirement: e.g. FAA/EASA requiring a specific Delegated Engineering Representative signature on material substitution and allowables data packages, or NRC 10 CFR 50 Appendix B / ASME BPVC Section III revisions naming a responsible metallurgical engineer for weld and heat-treat qualification records. Also state PE-board moves to close the industrial exemption (Ohio and NC have seen bills touching it) would put a licensed, personally liable signature on materials certifications.
Expansion of product-liability and recall exposure that names the materials disposition decision: if MDR/expert-witness practice or NHTSA/FDA recall investigations increasingly deposition the engineer who signed the material review board (MRB) disposition on a nonconforming lot, the role's ownership of ambiguous accept/reject calls becomes formally consequential.
If additive manufacturing qualification grows as a share of the work, per-build powder lot verification, witness-coupon testing and in-situ anomaly adjudication push more of the week onto the shop floor and into the lab rather than off it.
The limit. No plausible route to a meaningful trust premium: the buyer is an internal program manager or OEM customer who pays for a qualified certificate, not for a human author. Realistic combined ceiling is roughly the high-50s/low-60s, and it depends almost entirely on qualification-regime signature rules in aerospace, nuclear and medical devices — sectors that employ a minority of the 22,770.
| Los Angeles-Long Beach-Anaheim, CA | 1,250 | $125,200 +11% |
| Boston-Cambridge-Newton, MA-NH | 1,020 | $111,480 -1% |
| Seattle-Tacoma-Bellevue, WA | 820 | $156,130 +38% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 660 | $140,640 +25% |
| Houston-Pasadena-The Woodlands, TX | 560 | $123,250 +9% |
| San Jose-Sunnyvale-Santa Clara, CA | 550 | $156,730 +39% |
| Atlanta-Sandy Springs-Roswell, GA | 530 | $101,100 -10% |
| Denver-Aurora-Centennial, CO | 480 | $125,750 +11% |
| San Jose-Sunnyvale-Santa Clara, CA | 550 | $156,730 +39% |
| Seattle-Tacoma-Bellevue, WA | 820 | $156,130 +38% |
| Boulder, CO | 210 | $151,940 +35% |
We have no reported case of a named organisation automating this occupation. Not one deployment, not one announcement.
That is worth saying out loud next to a score of 44. The verdict above is about what the work exposes — what current AI could do to these tasks. It is not a claim that anyone has done it. For this occupation those two things have come apart completely: the capability argument is on this page, and the evidence column is empty.
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.