Why most of this work stays at the bench
Gem and diamond work is judgment plus hand control on a one-off object. Every piece of rough is different. The cutter reads the stone’s inclusions, grain and shape, then decides where to cleave or saw to save weight and avoid a fracture. That decision is made once, and it cannot be undone. Software can measure the stone, but someone still has to accept the trade-off between size, clarity and sparkle, and answer to the client who pays for it.
The hands matter as much as the eyes. Grinding and polishing facets against a lap wheel means feeling pressure, heat and the stone’s reaction through a dop stick. Setting and repairing stones in finished pieces means working in millimeters around soft metal and old settings. That blend of fine touch and irreversible choices is why the question “will ai replace diamond workers” has a slow answer rather than a dramatic one.
The pressure on the trade is real, but it mostly comes from elsewhere: a shrinking US workforce and cheaper cutting centers abroad. The Bureau of Labor Statistics counted about 22,440 workers in this occupation and projects a 3.2% decline from 2025 to 2035, with median pay of $52,540 (BLS, 2025). Fewer openings, not absent work, is the honest story.
What machines do, what they assist with, and what people keep
Software handles very little of the day on its own. The share of task time that runs without a person is 0%. That slice sits in the measuring and classifying steps: optical scanners that map proportions, symmetry and angles on a cut stone, and sorting rough by size and color before anyone touches a wheel.
Assisted work is where the newer tools land, at 6% of task time. Planning software models possible cuts from a 3D scan of the rough and shows the yield for each. Imaging systems flag inclusions and help document a stone before and after work. The machine proposes; the worker approves, adjusts and takes the blame if the stone chips.
The rest belongs to people, at 94% of task time. Cleaving and sawing by eye and feel, polishing facets to a finish a buyer will inspect under a loupe, mounting and tightening stones in a customer’s ring, and explaining to that customer why a stone was cut the way it was. Overall machine coverage of the job reads 9 out of 100, and you can see how that figure is built on the coverage method page.
How strong is the evidence?
Weak, and we say so. The quality parity grade for this job is D. At that grade there is no published head-to-head test of an automated system against a trained cutter or grader on this job’s real tasks, so we publish no parity number at all. Grading reports and sorting lines are widely automated in the industry, but industry adoption is not a measured comparison.
What would settle it: a blind study where graders and software assess the same parcel of stones against an agreed standard, and a yield study comparing planned-and-machine-cut stones with hand-planned cuts on matched rough. Until something like that is published, treat the score as a read of the task mix rather than a tested result. Our scoring method explains how grades move when better evidence arrives.
When the balance could shift
Most likely after 2046 (8 in 10 of our scenarios). The replacement-year method explains what that window is measuring.
Two things could pull it earlier. First, lab-grown supply makes rough far more uniform, and uniform input is what fixed automation is good at; a predictable stone needs fewer judgment calls. Second, grading and sorting are already the most automated steps, and once a factory buys the line, the marginal cost of running it is small compared with staffing a bench.
Two things hold it back. Most of the job is physical, and the robotics that fit it are fixed automation built for one repeated motion, not a flexible machine that can pick up an unfamiliar stone and decide where to cut. And the work is scattered across small shops, repair counters and custom jobs, where runs are short and capital equipment never pays back. Specialized hardware for a workforce of roughly 22,000 people (BLS, 2025) is a thin market.
What to do: keep a record of the hard saves and repairs you handle, because the tasks a machine cannot plan are the ones that keep you hired.
Staying needed in the trade
Lean into the work the task list leaves with people. Cutting and polishing decisions on irregular or included rough, where the yield call is yours. Setting and repair on finished jewelry, especially older pieces and heirlooms with worn mountings. And direct work with customers and designers, where you explain options, show the stone and set expectations before the saw comes out.
Two skills raise your floor. Learn to drive the planning and scanning software well, so you are the person who checks and overrides it rather than the one it replaces at a sorting table. And build gemological knowledge of lab-grown versus mined material and how to identify treatments, since the identification and disclosure side of the business is growing while cutting volume shrinks.
Nearby work is worth a look if you want more of the same skills in a different setting: jewelers and precious stone and metal workers, grinding and polishing workers, hand, and etchers and engravers. You can also see how the whole group sits in other production occupations and across manufacturing.
Want context? Put this job next to another in our side-by-side comparison, or check where hands-on trades land in the jobs that mostly need a person list.