EXPOSED
The screen half of this job — reading temperature and chemistry data, adjusting fuel/oxygen setpoints, logging heats, timing alloy additions — is exactly what modern process-control systems and model-predictive control already do better than a human eye on a gauge. The physical half — charging scrap, tapping and pouring molten metal, pulling samples, clearing slag and skulls, spotting a refractory failure or a bad tap before it becomes an injury — is hot, variable, and hostile to robots, which is why headcount shrinks per furnace rather than going to zero. The real pressure here is capital modernization and offshoring of smelting capacity, not chatbots.
Roughly flat across the period, with year-to-year wobble.
Median pay $42,250 → $54,430 +3.1% 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
-2.3%
Percentage only. The projection counts a different population from the 16,780 above — it includes self-employed workers, which for this occupation is most of them, so the two headcounts are not comparable.
Exposed, and shrinking
Both signals point the same way: the tasks are largely automatable and the BLS projects -2.3% by 2034. This is the case where the score and the forecast agree, and it is the one worth taking seriously.
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.
~2,000 openings a year on average, including replacing people who leave.
PolerBurnerMelterRodmanTopperChargerFettlerLevelerPlungerPuddlerPuncherRabblerRefinerRoasterSkimmerSlaggerSmelterStirrerRemelterConverterOre DryerGas TenderHot BallerKiln Firer
Holding it up: embodiment . Weakest point: trust premium .
Mixed — a routine tier and a judgment tier Level-3 automation already runs the burner curves, oxygen lancing schedules and heat logging on modern EAFs and reverb furnaces, but somebody still has to lance a frozen taphole, rod out a plugged spout, sledge skulls off the launder and read whether a hot spot on the shell means a refractory wear-through — the split is close to even, which is why this sits at 11 and not at 5 or 15.
Hands-on in uncontrolled environments You work within feet of 1,500°C metal in leathers and a face shield, hand-charging scrap and flux, dipping a sampling spoon into the bath, tapping and skimming — an environment with splash, radiant heat, gas leaks and no fixed geometry, which is why this is 17 rather than a 12 like a machine tender at ambient temperature.
No licence, no signature requirement No state licence gates a furnace tender; the pressure-vessel and emissions permits sit with the plant, EPA Subpart and OSHA 1910.269-type citations land on the employer, and your qualification is an in-house burner card or crane/hot-work sign-off that a new hire can earn in weeks — hence 3, not zero only because that sign-off exists.
Meaningful discretion You make live calls — hold the heat for another chemistry check, abort a tap on a suspect ladle, decide the slag is too cold to skim — but each is bounded by a written heat practice, a spec window and a shift supervisor within radio range, so it's real discretion with a procedure to fall back on, which is 9 and not 14.
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 (3/20) is whether the law requires a licensed human to sign. Trust premium (3/20) is whether buyers specifically pay for a person. Judgment and accountability (9/20) 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 15 of this occupation's 43 points (35%).
Embodiment (17/20) 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.
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 55/100, still EXPOSED.
As model-predictive control takes over setpoint tuning and heat logging, the residual job concentrates in the non-instrumented tier: judging refractory wear by eye, diagnosing off-spec heats when sensors disagree, hand-sampling and reacting to unexpected scrap chemistry. Task-mix shift on an existing two-tier job, no new law needed.
OSHA or state adoption of a molten-metal-specific standard requiring a named, qualified furnace operator to be present and to authorize each tap/pour — analogous to the certified operator requirements already in ASME boiler codes and to EPA Title V requirement for certified operators at some smelters. Insurer-driven versions exist: molten metal property carriers (FM Global data sheet 7-93) can condition coverage on trained-operator presence, which could harden into a named-signer requirement.
If a fatal breakout or water-into-molten-metal explosion produces litigation or an OSHA citation that turns on the operator's judgment to proceed, plant procedures typically respond by making stop-work and tap-authorization authority explicit and personally attributed. Watch for stop-work authority language in USW steel/aluminum master agreements and post-incident MOUs.
Already near ceiling. Would only move if furnace-tapping and slag-clearing environments got MORE variable — e.g. a shift to secondary/recycled feedstock (EV battery, e-scrap, mixed scrap streams) where charge composition is unpredictable and manual sampling/intervention per heat increases, as at Redwood Materials-type facilities.
The limit. Ceiling is low regardless. There is no plausible trust premium — buyers of copper cathode or billet specify chemistry, not who tended the furnace. And the dominant risk is not AI at all: capital modernization (one control room per several furnaces) and offshoring of primary smelting cut headcount even where every dimension above rises. A higher liability_shield protects the role's legal necessity, not the number of them.
| Chicago-Naperville-Elgin, IL-IN | 700 | $63,170 +16% |
| Cleveland, OH | 520 | $65,660 +21% |
| Birmingham, AL | 420 | $56,250 +3% |
| Tampa-St. Petersburg-Clearwater, FL | 370 | $53,130 -2% |
| Pittsburgh, PA | 320 | $60,810 +12% |
| Charlotte-Concord-Gastonia, NC-SC | 240 | $61,780 +14% |
| Spokane-Spokane Valley, WA | 240 | $85,370 +57% |
| Reading, PA | 210 | $68,990 +27% |
| Spokane-Spokane Valley, WA | 240 | $85,370 +57% |
| Utica-Rome, NY | 190 | $77,560 +42% |
| Portland-Vancouver-Hillsboro, OR-WA | 190 | $74,060 +36% |
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 43. 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.