EXPOSED
The core loop — weigh and load ingredients, set mixer speed and time, watch gauges, log batch records, pull samples — is exactly what process control systems and recipe automation have been eating for decades, and AI-assisted batch optimization accelerates it. What holds is the physical side: hauling drums and sacks, hooking up hoses, clearing jammed or caked material, breaking down and sanitizing tanks and blades between runs, and catching an off-viscosity or off-color batch by feel and smell before the sensor flags it. No license protects the role, buyers never meet the operator, and discretion is real but bounded by the recipe sheet.
Part 2020 shock, part continued decline in the years since.
Median pay $37,780 → $48,990 +3.7% 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
-6.8% 101,100 → 94,300 on the projections basis
Exposed, and shrinking
Both signals point the same way: the tasks are largely automatable and the BLS projects -6.8% 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.
~8,800 openings a year on average, including replacing people who leave.
MixerPuggerPulperSlakerBatcherBlenderBlungerChurnerReducerThinnerCrutcherDip DyerOperatorBoss DyerDry MixerDye MakerDye MixerGum MixerInk MakerInk MixerMarinatorMud MixerOil MixerOre Mixer
Holding it up: embodiment . Weakest point: trust premium .
Mixed — a routine tier and a judgment tier Recipe download, mass-flow metering, timed agitation and automatic batch logging already run unattended in modern PLC-controlled plants, so the setting and tending half of the job is largely solved — the 9 rather than a 4 comes from the manual charging of 50-lb sacks and drums, hose changeovers, and reacting to material that cakes, bridges in the hopper, or won't disperse.
Hands-on in uncontrolled environments You work at the vessel: climbing platforms to charge from above, wrestling totes and drum pumps, scraping caked residue off blades and baffles, breaking down and CIP-ing tanks between runs, in dust, heat, solvent vapor and confined-space entry — a 15 rather than 18 because it is one fixed plant floor with known equipment, not an unpredictable outdoor site.
No licence, no signature requirement Nothing in the job requires a credential; you may carry a forklift cert, HAZMAT or confined-space training and follow the plant's GMP or FDA/OSHA process safety procedures, but the batch record is signed off by QC and the process engineer, and the 3 reflects that any liability for a bad lot lands on the company, not on you personally.
Meaningful discretion You decide when a blend looks right, whether to add extra mix time, when to stop the line and call QC on an off-color or off-viscosity lot — real calls with scrap cost attached, but the formulation, tolerances and hold criteria are written on the batch sheet by someone else, which caps this at 8 rather than 13.
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 (9/20 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 (8/20) 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 14 of this occupation's 38 points (37%).
Embodiment (15/20) 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 54/100, still EXPOSED.
Task-mix shift is genuine here: this job has a routine tier (weigh, load, set time/speed, log) and a judgment tier (diagnosing why a batch went off-spec, caked material, blade wear, ingredient lot variability, wet-vs-dry season behavior of powders). If recipe automation and inline NIR/rheometry take the routine tier, the surviving role is troubleshooting and lot-variability compensation across many lines rather than tending one. The count of workers falls while the residual work resists.
Cannabis and infant-formula blending are the tightest cases: state cannabis rules (e.g. Colorado MED, California DCC) already require a named batch operator on the manifest, and the 2023 infant formula recalls drove FDA toward operator-level attestation on nutrient premix addition. Expansion of named-operator manifest requirements into more state-regulated ingredient categories raises this.
FDA's Food Traceability Rule (FSMA 204, compliance date extended to 2028) and DSCSA-style batch-record integrity expectations push toward a named, trained human attesting to each batch record. In pharma and dietary supplement blending, cGMP (21 CFR 211.188 / 111.260) already requires a signed, dated batch production record entry by the person performing weighing and mixing plus a second-person verification of component additions — if plants respond to AI-generated records by formally designating the operator as the attesting signer rather than a QA manager, the shield attaches to the role. Watch for FDA warning letters citing unverified automated component addition.
If plants adopt operator authority to hold or reject a batch — a documented stop-batch right analogous to the andon cord, tied to cGMP deviation reporting — the role owns a consequential call rather than executing a recipe sheet. Some union contracts in food processing (UFCW, BCTGM) already contain food-safety stop-work language; broader adoption would move this.
Not a scoring lever so much as a floor: sanitizing tanks and blades between allergen or flavor changeovers, confined-space tank entry, hose hookups, and clearing caked material remain outside robot reach. Allergen-changeover cleaning validation requirements (FDA FALCPA, and the 2021 sesame addition) increase the share of the shift spent on non-automatable wet cleaning and swab verification.
The limit. Trust premium has no realistic route: buyers of blended pigment, resin, dough, or premix never learn who ran the mixer and there is no consumer-facing artisan story to sell, unlike small-batch brewing or perfumery which sit under different SOC codes. Even with all levers, the shields available here are institutional (cGMP signatures, state manifests) rather than licensure — they follow the plant, not the person, so they can be reassigned to a QA technician at any time. Realistic ceiling is mid-50s, and it comes with a materially smaller headcount.
| Chicago-Naperville-Elgin, IL-IN | 6,550 | $50,230 +3% |
| New York-Newark-Jersey City, NY-NJ | 3,430 | $56,780 +16% |
| Los Angeles-Long Beach-Anaheim, CA | 2,080 | $48,870 +0% |
| Dallas-Fort Worth-Arlington, TX | 2,070 | $45,320 -7% |
| Houston-Pasadena-The Woodlands, TX | 1,550 | $43,820 -11% |
| Atlanta-Sandy Springs-Roswell, GA | 1,470 | $47,180 -4% |
| Cleveland, OH | 1,190 | $48,350 -1% |
| Phoenix-Mesa-Chandler, AZ | 1,160 | $48,500 -1% |
| Sacramento-Roseville-Folsom, CA | 310 | $75,790 +55% |
| Duluth, MN-WI | 120 | $73,210 +49% |
| Wausau, WI | 190 | $69,080 +41% |
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 38. 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.