Why glasses still get fitted face to face
Will AI replace opticians? The honest answer sits in the task mix. A dispensing optician measures pupillary distance and vertex distance against a real face, then adjusts temples, nose pads and frame angle until the lenses sit where the prescription assumes they sit. Software can calculate those numbers. It cannot heat a plastic frame and bend it to fit one head.
The second reason is judgment about people, not optics. A customer walks in with a new progressive prescription, a narrow bridge and a budget. Somebody has to translate the prescription into a lens choice, a frame that holds it, and an honest explanation of what the first week will feel like. When the glasses come back wrong, somebody has to work out whether the fault is the lens order, the fit, or the prescription itself, and then handle the remake.
Scale matters too. About 73,530 people work as dispensing opticians in the US, with median pay of $47,260 a year, and employment is projected to grow 4.2% between 2025 and 2035 (BLS, 2025). That is steady work attached to a physical product, which is why our model puts the hardware needed to copy the hands-on part in the dexterous humanoid class rather than in software. How each of those questions is scored is set out in our scoring methodology.
What AI does, what it helps with, and what stays with people
Digital tools already handle part of the measurement and paperwork. Tablet-based centration systems capture frame and pupil measurements from photos, practice software writes the lab work order, and record-keeping and billing run largely on their own. That group covers 0% of task time in our estimate.
A larger overlap is assistance. Frame recommendation from face shape and prescription strength, lens option comparisons, and checks on whether a measurement looks out of range are all faster with software, but an optician signs off on them. Assisted tasks account for 35% of the job’s task time.
The rest is the dispensary floor: taking the measurements, fitting and adjusting frames, repairing broken hinges and pads, and teaching a first-time wearer how to insert, remove and care for contact lenses. That group stands at 65% of task time, which is why the overall answer to “Can AI do it?” reads 19 out of 100 on our coverage scale.
What the evidence actually shows
Most published AI work in eye care sits one door down from the dispensary. Image-reading models for retinal disease have been tested hard; automated frame fitting and lens dispensing have not. For dispensing opticians, the evidence grade on “Is it better than a person?” is D, which on our scale means there is no direct head-to-head test of AI against trained opticians doing this job’s tasks. We give no parity number where nothing has been measured, and the quality parity method explains why.
A few specific studies would settle it. Measured agreement between app-based centration and a certified optician’s measurements, across many face shapes and frame styles. Remake and return rates for glasses dispensed with and without a human fitting. Comfort and adaptation outcomes for progressive lenses ordered online versus fitted in person. Until results like those are published and repeated, claims in either direction are opinion.
When the work could shift
Most likely after 2042 (8 in 10 of our scenarios). What that window counts, and how it is built, is described on the replacement-year page.
Two things could pull it earlier. Phone-based measurement and virtual try-on keep improving, and the software side is cheap next to a trained dispensary. Online and direct-to-consumer eyewear also moves volume out of the store entirely, which thins entry-level dispensing roles before it touches senior ones.
Two things hold it back. The adjusting, repairing and refitting work needs hands with fine control near a customer’s eyes, and that hardware is neither cheap nor common. Licensing and certification rules in many states also decide who may take measurements and dispense, so a tool can assist without being allowed to sign off.
Good to know: the pressure here shows up first as fewer trainee dispensing hours in high-volume retail, not as stores without opticians.
How to stay needed as a dispensing optician
Lean into the parts of the job that a camera cannot finish. Complex fitting work, including progressives, high-index lenses, children’s frames and low-vision aids. Repairs and refits, which bring customers back and build the relationship that online ordering lacks. Contact lens instruction, where a nervous first-timer needs patience and a second attempt.
Two skills raise your value rather than compete with the tools. Learn the digital measurement and lens-design software well enough to spot when its output is wrong. Then get better at troubleshooting: reading a failed prescription or a bad adaptation and explaining the fix in plain words.
If you want to move sideways, the closest work is clinical rather than retail. Compare the task mix with optometrists, ophthalmic medical technicians and orthoptists. You can also see where this job sits among health technologists and technicians, read the wider healthcare sector view, or put two of these roles side by side. For context on hands-on work in general, the list of jobs that mostly need a person is a useful starting point.