noworkleft.com

Home › Jobs › Helpers--Roofers

Construction and Extraction · SOC 47-3016

Helpers--Roofers

Help roofers by performing duties requiring less skill. Duties include using, supplying, or holding materials or tools, and cleaning work area and equipment.

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 18 tasks O*NET lists for this occupation:

Could an LLM do it?

0%

of tasks: an LLM alone could do the task or cut its time by half. Human raters say 0%, GPT-4 says 0%.

Counting tasks that need extra software around the model: 0%–0%.

Is it happening?

0 of 18 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?

18 of 18 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
Check to ensure that completed roofs are watertight. no rater none seen
Set ladders, scaffolds, and hoists in place for taking supplies to roofs. no rater none seen
Clean work areas and equipment. no rater none seen
Maintain tools and equipment. no rater none seen
Provide assistance to skilled roofers installing and repairing roofs, flashings, and surfaces. no rater none seen
Locate worn or torn areas in roofs. no rater none seen
Sweep and clean roofs to prepare them for the application of new roofing materials. no rater none seen
Remove old roofing materials. no rater none seen
Perform emergency leak repairs and general maintenance for a variety of roof types. no rater none seen
Unload materials and tools from work trucks, and unroll roofing as directed. no rater none seen
Hoist tar and roofing materials to roofs, using ropes and pulleys, or carry materials up ladders. no rater none seen
Clear drains and downspouts and clean gutters. no rater none seen
Cover roofs with layers of roofing felt or asphalt strips before installing tile, slate, or composition materials. no rater none seen
Attach sheets of metal to roof boards or building frameworks when installing metal roofs. no rater none seen
Apply shingles, gravel, or asphalt over the top layer of tar to protect the roofing material. no rater none seen
Chop tar into small pieces, and heat chopped tar in kettles. no rater none seen
Attach roofing paper and composition shingles, using nails. no rater none seen
Place tiles, nail them to roof boards, and cover nailheads with roofing cement. (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.