noworkleft.com

Home › Jobs › Carpenters

Construction and Extraction · SOC 47-2031

Carpenters

Construct, erect, install, or repair structures and fixtures made of wood and comparable materials, such as concrete forms; building frameworks, including partitions, joists, studding, and rafters; and wood stairways, window and door frames, and hardwood floors. May also install cabinets, siding, drywall, and batt or roll insulation. Includes brattice builders who build doors or brattices (ventilation walls or partitions) in underground passageways.

Will AI replace this job?

The data cannot say, and anyone who gives you a percentage for “replaced” is guessing. What it does show is a gap between three things, for the 29 tasks O*NET lists for this occupation:

Could an LLM do it?

3%

of tasks: an LLM alone could do the task or cut its time by half. Human raters say 3%, GPT-4 says 3%.

Counting tasks that need extra software around the model: 14%–24%.

Is it happening?

0 of 29 tasks

show measurable use in Claude conversations (Anthropic Economic Index). Anthropic’s occupation score, which counts only tasks that were also rated feasible, is 0%. Observed is a subset of “could” by construction.

Claude only, mostly US users, sampled periods in 2025. Not all AI use, and not employment.

What is left?

22 of 29 tasks

were rated “no LLM speed-up” by every rater (so, by Anthropic’s method, they cannot show up as observed use either). That is the part of the job these studies see no sign of AI touching.

It does not mean those tasks are safe. It means these methods found nothing there.

Task by task

Sorted with the tasks seen in use first. The three coloured squares are the three raters, in this order: human raters, GPT-4, GPT-4 with an alternative rubric. “Usage score” is Anthropic’s published task_penetration value; 0 means none seen.

LLM aloneneeds extra softwareno speed-up
Task (O*NET 31.0)Could an LLM?Claude usage score
Arrange for subcontractors to deal with special areas, such as heating or electrical wiring work. all raters none seen
Maintain job records and schedule work crew. all raters none seen
Prepare cost estimates for clients or employers. all raters none seen
Maintain records, document actions, and present written progress reports. all raters none seen
Select and order lumber or other required materials. 2 of 3 raters none seen
Inspect ceiling or floor tile, wall coverings, siding, glass, or woodwork to detect broken or damaged structures. 1 of 3 raters none seen
Study specifications in blueprints, sketches, or building plans to prepare project layout and determine dimensions and materials required. 1 of 3 raters none seen
Measure and mark cutting lines on materials, using a ruler, pencil, chalk, and marking gauge. no rater none seen
Follow established safety rules and regulations and maintain a safe and clean environment. no rater none seen
Shape or cut materials to specified measurements, using hand tools, machines, or power saws. no rater none seen
Install structures or fixtures, such as windows, frames, floorings, trim, or hardware, using carpenters' hand or power tools. no rater none seen
Verify trueness of structure, using plumb bob and level. no rater none seen
Build or repair cabinets, doors, frameworks, floors, or other wooden fixtures used in buildings, using woodworking machines, carpenter's hand tools, or power tools. no rater none seen
Remove damaged or defective parts or sections of structures and repair or replace, using hand tools. no rater none seen
Install rough door and window frames, subflooring, fixtures, or temporary supports in structures undergoing construction or repair. no rater none seen
Erect scaffolding or ladders for assembling structures above ground level. no rater none seen
Examine structural timbers and supports to detect decay, and replace timbers as required, using hand tools, nuts, and bolts. no rater none seen
Fill cracks or other defects in plaster or plasterboard and sand patch, using patching plaster, trowel, and sanding tool. no rater none seen
Dig or direct digging of post holes and set poles to support structures. no rater none seen
Finish surfaces of woodwork or wallboard in houses or buildings, using paint, hand tools, or paneling. no rater none seen
Cover subfloors with building paper to keep out moisture and lay hardwood, parquet, or wood-strip-block floors by nailing floors to subfloor or cementing them to mastic or asphalt base. no rater none seen
Bore boltholes in timber, masonry or concrete walls, using power drill. no rater none seen
Assemble and fasten materials to make frameworks or props, using hand tools and wood screws, nails, dowel pins, or glue. no rater none seen
Perform minor plumbing, welding, or concrete mixing work. no rater none seen
Anchor and brace forms and other structures in place, using nails, bolts, anchor rods, steel cables, planks, wedges, and timbers. no rater none seen
Work with or remove hazardous material. (supplemental) no rater none seen
Build sleds from logs and timbers for use in hauling camp buildings and machinery through wooded areas. (supplemental) no rater none seen
Construct forms or chutes for pouring concrete. (supplemental) no rater none seen
Apply shock-absorbing, sound-deadening, or decorative paneling to ceilings or walls. (supplemental) no rater none seen
How to read this. These are averages over task lists written for the US labour market, rated by people and by GPT-4 using early-2023 model capabilities, next to a count of what one company’s chatbot users actually ask for. A task being “exposed” says nothing about whether employers will cut jobs, shift work, or hire more of you. See how to read an exposure score.

Try it with your own task list

Sources: task list from O*NET® Database (U.S. Department of Labor, CC BY 4.0; modified by noworkleft.com); “could” labels from Eloundou et al., GPTs are GPTs (OpenAI/University of Pennsylvania, MIT licence, labels from early-2023 models); “seen in use” from the Anthropic Economic Index (Claude usage only, CC BY). Dates, versions and limits.

Other jobs in Construction and Extraction

Data built 2026-11-06. O*NET 31.0. Task match rates: 99.4% with the OpenAI/Penn labels, 99.2% with Anthropic usage.