■ CRITICAL RISK ■ Personal Services
The market replaced this job before AI got the chance — nobody repairs a $12 quartz clock. What survives is a tiny, aging craft trade serving antiques and heirlooms, which is less a career than a calling with a waiting list.
“Smartphones replaced clocks. And the people who fix them.”
Our AI replacement risk score — how we score jobs
Traditional clock repair is watchmaker's work at a larger scale: disassembling a mechanical movement, cleaning decades of dried oil from pivots and bushings, replacing worn parts (often by fabricating them on a lathe, since the manufacturer dissolved in 1964), regulating the escapement, and reassembling several hundred parts in the right order. It demands fine motor skill, horological knowledge, and the patience of a monk. None of that was automated — it was obsoleted. Quartz movements made timekeeping so cheap that repair stopped making economic sense for anything ordinary, and phones then made standalone clocks decorative.
So the automation story here is unusual: AI barely features. There's no robot rebuilding grandfather clock movements, and there won't be — the volumes are microscopic and every job is different. What technology did instead was destroy the demand side. The repair that remains serves antique and sentimental pieces: grandfather clocks, heirloom mantel clocks, tower clocks on courthouses, and high-end mechanical pieces whose owners pay for craftsmanship, not timekeeping. That niche is stable, even underserved — established repairers commonly report backlogs of months, because the trade's practitioners are retiring faster than new ones train.
Our 92 score reflects the occupation's arc, not robot competition: employment collapsed by orders of magnitude and will never return, and no one should read 'backlog at the antique shop' as a growth industry. But it's worth being precise about the paradox — for the few hundred people who master it, this is one of the more automation-proof ways to earn a living we cover, precisely because the machines already took everything they wanted decades ago.
Automatability: our editorial assessment of current and near-term AI capability
The disruption already happened — quartz in the 1970s-80s and smartphones after 2007 did the demolition, and there's nothing left for AI to take. What's happening this decade is demographic: the remaining repairers are aging out, demand from antique owners persists, and waitlists lengthen. The occupation keeps shrinking in headcount while, strangely, becoming more secure for each individual practitioner still standing.
In the antique and heirloom niche, genuinely yes — surviving repairers often carry months-long backlogs because demand from grandfather clock and heirloom owners outstrips the shrinking supply of qualified people. Ordinary clock repair is dead; nobody fixes a $12 quartz wall clock. The living is made on craftsmanship-priced restoration, not volume.
No — quartz movements and then smartphones did, decades before modern AI existed. Cheap electronic timekeeping made repair uneconomical for everyday clocks, collapsing the trade's volume. AI is essentially irrelevant to what remains: no algorithm rebuilds a 200-year-old mechanical movement, and the surviving niche is protected by exactly that fact.
Only with clear eyes: it's a micro-trade, not a career track with job postings. That said, the few who train seriously — often through horological schools or apprenticeships shared with watchmaking — enter a market where retiring experts leave backlogs behind. Pair it with watch repair for a broader clientele. It suits the self-employed craftsperson temperament, not someone seeking employment.
Because supply is collapsing faster than demand. Antique clocks keep existing and keep needing service every decade or so, but the repairers who learned the trade in its heyday are retiring with few successors. The result is the odd economics of dying crafts: tiny total employment, yet real scarcity pricing for the remaining skilled hands.