Millwrights
In short: today's best AI models could do about 0% of this job's working time (projected from METR's April 2026 measurement and Epoch's November 2026 index) if the work were set up for them, and about the same by the end of 2028 if progress keeps its long-run pace. The people and hands-on work, about 100% of the time, stays with people.
Install, dismantle, or move machinery and heavy equipment according to layout plans, blueprints, or other drawings.
40,330 people do this job in the US. $65,700 a year.
A typical 40-hour week
Not yet reviewed task by task- AI could take on0 h
- A person checks AI's work0 h
- Desk work AI can't do yet0 h
- Stays with people40 h
Millwrights, now: of a 40-hour week, AI could take on 0 hours, 0 hours become checking its work, 0 hours are desk work it can't do yet, and 40 hours stay with people.
Estimates for a typical week of this job, if progress keeps its long-run pace, counting work the best models finish four times in five. Every task is below.
of the working time. Now is projected from METR's April 2026 measurement and Epoch's November 2026 index.
- 0%
- 0%
- 100%
of the work: Anthropic's measure of what people already bring to Claude at work (2025). It counts help with part of a task, so it can be higher than the share within reach.
On the chosen pace and reliability. Estimate.
Share of the working time within reach, now to 2030
Change assumptionsHide assumptionsLong-run pace, four in five
What each choice means: , , , , . Every pace starts from .
At this date and pace, models finish tasks up to 2.4 work days long four times in five.
How far this has come
With GPT-4, measured in March 2023, 0% of this job's working time was within reach at . With Claude 3.7 Sonnet, measured in February 2025, 0%. With Claude Mythos Preview (early), measured in April 2026, 0%. Today's estimate: 0%.
A backcast: today's task data with the best model METR had measured by each date. It shows how fast the models have moved, not what anyone forecast at the time.
Measured use over time
Of the Claude.ai conversations Anthropic sampled and tied to a work task, 0.042% went to this job's tasks in May 2026, the 244th largest of the 447 jobs that showed up at all; 1 of its 23 tasks showed up. Within each of Anthropic's methods, never across a change: 0.003% to 0.0043% (released February and March 2025); 0.034% to 0.042% (April to May 2026).
Each release
| Conversations | Share | Rank | Tasks seen |
|---|---|---|---|
| Released February 2025 | 0.003% | 655 of 708 | 1 of 23 |
| Released March 2025 | 0.0043% | 597 of 703 | 1 of 23 |
| The week of Aug 4, 2025 (released September 2025) | none | n/a | 0 of 23 |
| The week of Nov 13, 2025 (released January 2026) | none | n/a | 0 of 23 |
| The week of Feb 5, 2026 (released March 2026) | none | n/a | 0 of 23 |
| April 2026 (released June 2026) | 0.034% | 269 of 434 | 1 of 23 |
| May 2026 (released June 2026) | 0.042% | 244 of 447 | 1 of 23 |
Measured, not estimated: Anthropic's samples of Claude.ai conversations, each tied by a model to an O*NET task, from the Anthropic Economic Index. It says where Claude's use goes, not how much of this job's working time it covers, so it is not on the same scale as the reach above. The releases differ in their sample, their classifier and the smallest share they report (the June 2026 file rounds each task to a hundredth of a percent, so a smaller task reads as none), which is why a change of method starts a new line. At least 96% of each older release's task-tied use matched this site's tasks, and all of the newest. Claude.ai only, so use through other tools and other companies' models is not in it. Every job ranked.
Robots and the physical work
Anthropic rates 100% of this job's working time as physical work. For 66% of the working time, a robot has been shown doing most of the task somewhere: 66% in a purpose-built cell. 0.66 of 3, the 268th highest of the 923 jobs.
- No robot shown 34%
- In a purpose-built cell 66%
Anthropic's rating, not this site's estimate: Claude Opus 5, searching the web, rated each physical O*NET task by the kind of setting in which a robot it could cite (deployed, sold or demonstrated) does most of the task, weighted by its own estimate of each task's share of the time (What work can robots do?, September 30, 2026, data CC BY 4.0; every job ranked). A robot shown somewhere is not a robot in your building: cost, setup and safety decide that. The share within reach on this page still counts no body work. The robot behind each rated task is under the task below.
The tasks
Sorted by . 23 tasks from .
| Task | Kind | |||
|---|---|---|---|---|
| Assemble and install equipment, using hand tools and power tools.Done daily by most who do it. Robots: in a purpose-built cell, KUKA robotic engine-block assembly line (Scania powertrain) (deployed). Anthropic's rating. | 9% | 5 h1.7 h to 2.1 work days | Stays with people | |
| Assemble machines, and bolt, weld, rivet, or otherwise fasten them to foundation or other structures, using hand tools and power tools.Done daily by most who do it. Robots: in a purpose-built cell, Dürr flex decking / flex bolt marriage and automatic bolting stations (deployed). Anthropic's rating. | 9% | 6.7 h2.5 h to 2.5 work days | Stays with people | |
| Dismantle machinery and equipment for shipment to installation site, performing installation and maintenance work as part of team.Done more than yearly by most who do it. Robots: no robot shown. Anthropic's rating. | 9% | 1.3 work days3.3 h to 4.2 work days | Stays with people | |
| Construct foundation for machines, using hand tools and building materials such as wood, cement, and steel.Done more than monthly by most who do it. Robots: in a purpose-built cell, Elematic Shuttering Robot E9 (automatic shuttering station) (sold). Anthropic's rating. | 8% | 1.3 work days3.3 h to 4.2 work days | Stays with people | |
| Attach moving parts and subassemblies to basic assembly unit, using hand tools and power tools.Done more than weekly by most who do it. Robots: in a purpose-built cell, Duerr flex decking marriage station (NEXT.assembly) (sold). Anthropic's rating. | 6% | 3.3 h50 min to 6.7 h | Stays with people | |
| Weld, repair, and fabricate equipment or machinery.Done daily by most who do it. Robots: in a purpose-built cell, MaxQ Repair (RAMLAB software and sensor suite on Techman/Panasonic/Hyundai robots) (sold). Anthropic's rating. | 6% | 5 h1.7 h to 1.3 work days | Stays with people | |
| Dismantle machines, using hammers, wrenches, crowbars, and other hand tools.Done daily by most who do it. Robots: no robot shown. Anthropic's rating. | 6% | 3.3 h50 min to 1.7 work days | Stays with people | |
| Replace defective parts of machine, or adjust clearances and alignment of moving parts.Done daily by most who do it. Robots: no robot shown. Anthropic's rating. | 5% | 2.5 h50 min to 6.7 h | Stays with people | |
| Position steel beams to support bedplates of machines and equipment, using blueprints and schematic drawings to determine work procedures.Done more than weekly by most who do it. Robots: in a purpose-built cell, Zeman SBA Conti / SBA steel beam assembler (Zeman Bauelemente, Lincoln Electric group) (sold). Anthropic's rating. | 4% | 5 h1.7 h to 1.7 work days | Stays with people | |
| Conduct preventative maintenance and repair, and lubricate machines and equipment.Done daily by most who do it. Robots: no robot shown. Anthropic's rating. | 4% | 2.5 h25 min to 6.7 h | Stays with people | |
| Move machinery and equipment, using hoists, dollies, rollers, and trucks.Done daily by most who do it. Robots: in a purpose-built cell, KUKA omniMove heavy-duty AGV at AMAG Ranshofen (deployed). Anthropic's rating. | 4% | 2.5 h25 min to 1 work day | Stays with people | |
| Fabricate and dismantle parts, equipment, and machines, using a cutting torch or other cutting equipment.Done daily by most who do it. Robots: no robot shown. Anthropic's rating. | 4% | 3.3 h1.3 h to 6.7 h | Stays with people |
Within reach, per task: the share of the task's time whose instances are short enough for a model to finish at the chosen date, pace and reliability. Half by: the date half of that time comes within reach.
Jobs that share skills with this one
From O*NET's related occupations. Within reach now and by the end of 2028, long-run pace, four in five.
- Aircraft Structure, Surfaces, Rigging, and Systems Assemblers4% now, 4% by 2028
- Maintenance and Repair Workers, General11% now, 13% by 2028
- Industrial Machinery Mechanics13% now, 13% by 2028
- Engine and Other Machine Assemblers9% now, 9% by 2028
- Boilermakers3% now, 3% by 2028
- Machinists27% now, 33% by 2028
Where this job works
The industry groups that employ the most of it (BLS, May 2025).
Where the most of these jobs are
The metro areas with the most people in it (BLS, May 2025).
Questions about this job
- How much of the work of Millwrights is within reach of AI now?
- An estimated 0% of the working time is within reach of AI models now and 0% by the end of 2028, on METR's long-run pace at four in five. Within reach means a model could finish work that long at that reliability. It is not a forecast of jobs lost or of savings.
- Which tasks of Millwrights come within reach first?
- No task has a meaningful share of its time within reach now or by the end of 2028, on the long-run pace at four in five: the work is mostly people work and body work, which the site does not count.
- Which tasks of Millwrights stay with people?
- Tasks that are people work or body work are not counted as within reach at any date. The largest: Assemble and install equipment, using hand tools and power tools; Assemble machines, and bolt, weld, rivet, or otherwise fasten them to foundation or other structures, using hand tools and power tools; and Dismantle machinery and equipment for shipment to installation site, performing installation and maintenance work as part of team. Together such tasks are about 100% of the working time.
- Do people already use AI for the work of Millwrights?
- In the Claude.ai conversations Anthropic sampled for May 2026 and tied to a work task, 0.042% went to this job's tasks, the 244th largest share of the 447 jobs that showed up; 1 of its 23 tasks appeared. Anthropic's labor market measure puts the share of the job's tasks seen in real Claude use at work at 0%. These are measures of use, not of how much of the work AI could finish, and they cover Claude only.
- Can robots do the physical work of Millwrights?
- Anthropic's robot exposure rating (September 2026) finds 100% of the working time is physical work, and that a robot has been shown doing most of the task, somewhere, for 66% of the working time. Its index is 0.66 of 3. One example: KUKA robotic engine-block assembly line (Scania powertrain) for "assemble and install equipment, using hand tools and power tools". A robot shown somewhere is not one in every workplace; this site's own estimates still count no physical work as within reach.
- How is this estimated?
- From the 23 tasks O*NET lists for the occupation, each with an estimated length checked against GDPval and a share of the working time, each classed as desk, people or body work, and METR's measurements of how long a task the best AI models can finish, carried forward to the date from the latest, Claude Mythos Preview (early) in April 2026. Data checked November 10, 2026.
This job in your state or metro area
Embed these figures on your site
A small card with the share of working time within reach now and at the end of 2027, 2028 and 2030, the date and a link back. Free under CC BY 4.0; the card carries the credit.
Your own version of this job
The page above is the occupation. Your job is your title and what you actually do: give both, and they are read against these tasks in your browser, nothing sent anywhere. The headline is free: how much of your working time is within reach now and by the end of 2028, and how many of the tasks your description names.
This occupation has not been reviewed task by task yet, so the full personal report is not offered for it; 402 of 923 occupations are reviewed so far, largest first.