The Forty Hours That AI Can't Replace
What can a mechanical watch teach brands about surviving the AI revolution? As competent output becomes abundant, Mark Armstrong, Executive Creative Director at Bulletproof APAC, writes that craftsmanship, human judgement and the invisible hours behind great design will become a brand's most valuable asset.
I've had a fascination with watches for as long as I can remember. It started with my grandad letting me wind his pocket watch. The sound it made, the feel of the action, and the complete lack of understanding of how that action could then keep time.
That curiosity turned, years later, into properly collecting watches. I remember the first time someone showed me a movement under a loupe and pointed out the anglage, the polished bevel along every edge of every bridge, cut by hand, at an angle a machine has no reason to bother with because nobody but a watchmaker and a collector will ever see it. That's roughly forty hours of invisible, unnecessary human labour in a single watch.
I've been thinking about that detail a lot lately, for reasons that have nothing to do with watches. For a hundred years, the story of progress has been the story of making things faster, cheaper, and more consistent. AI is about to finish that story. When the cost of producing something competent - a design, a product, a piece of content, a piece of code - collapses toward zero, competence stops being a differentiator and it becomes the baseline everyone stands on.
Which raises an uncomfortable question for brands. If AI can produce ‘good enough’ infinitely, for free, what's left to sell? Watchmaking answered this question fifty years before AI made it everyone else's problem.
It’s happened before
In the 1970s, quartz movements did to mechanical watchmaking exactly what AI now threatens to do to skilled work generally: more accurate, far cheaper, and requiring none of the years of hand-taught skill mechanical watchmaking demanded.
By the late 1970s quartz had overtaken mechanical watches in popularity outright, and an industry built for a century on hand-finished movements nearly collapsed. Famous houses went insolvent.
Then something counterintuitive happened. Once quartz had made accurate timekeeping universal and essentially free, a generation that grew up wearing quartz watches started looking at mechanical ones differently. Not as outdated technology, but as ‘craft objects’. Houses like A. Lange & Söhne, dormant for decades, were revived in the 1990s built unapologetically around hand-finishing and mechanical seriousness. Nobody needed a mechanical watch to know what time it was. That was never the point. It was an identity decision and consumer desire that turned into one of the more valuable identity decisions in the history of luxury goods.
Automation didn't kill craftsmanship. It clarified what craftsmanship was actually worth, by removing the one thing that used to disguise it, which was function.
What ‘worth more’ looks like
If you want to see where this ends up, look at the independent watchmakers. Small ateliers where a handful of people produce a few dozen or a few hundred watches a year, entirely by hand, and sign their own name to the dial. Rexhep Rexhepi trained at Patek Philippe's own watchmaking school before founding his own house, Akrivia. Today the waiting list for an Akrivia is longer than the waiting list for a steel Patek Nautilus, one of the most famously unobtainable watches in mass-luxury. One recent buyer was quoted an eleven-year wait for a watch from a small Japanese maker, and didn't blink.
None of this is about telling time better. A thirty-dollar quartz watch is objectively more accurate than any hand-finished mechanical watch ever made. The entire value sits in the parts nobody needs, the hand-bevelled angles that take years to learn, the fact that a real person touched every component before it left the workshop. The premium isn't for the object. It's for the evidence of unhurried human attention, applied somewhere nobody was forced to apply it.
It's happening again, and the tell is showing up early
We're not waiting a generation to see this pattern repeat with AI. In February, Gucci ran an AI-generated campaign at Milan Fashion Week and was met with a backlash accusing the house of abandoning the craftsmanship it built its name on.
A March 2026 Gartner survey found half of US consumers would rather do business with brands that avoid generative AI altogether. And Kering, one of the largest luxury conglomerates on earth, has recently opened a craftsmanship academy built to train up to 2,000 students a year. That's the watch-industry playbook exactly. Train the hands now, own the scarce asset later.
Throughout my career I’ve been fortunate to work with brands that already sit on this same asset, without having had to invent it under pressure. Johnnie Walker's entire flavour identity has passed through just eight Master Blenders in two hundred years. Penfolds' Grange only exists because Max Schubert kept making it in secret after the board ordered him to stop. Neither of those is a story an algorithm could produce, they're proof that the value was never in the process, it was in a specific person's judgment, held onto stubbornly, over generations.
Why scarcity works this way
The mechanism isn't nostalgia. It's supply and demand, applied to something we've never had to think about this way before, human attention itself.
Since the industrial revolution, skilled human output was the scarce input and machine output was the abundant one. AI inverts that, the way quartz once inverted it for time-telling. When competent output becomes infinite and nearly free, the thing that stays scarce, hours of trained human judgment, applied by hand and over time, doesn't lose value. It gains it, because scarcity is relative.
And there's a provenance angle too. As AI-generated output becomes the default, ‘made by a specific person, who spent real hours on this’ becomes a verifiable, differentiating claim. The same way ‘organic’ only became a meaningful label once industrial food made ‘just food’ the invisible default.
The strategic question brands need to answer now, not later
This divide is being drawn right now, and most brands don't seem to know it's happening. Having craft heritage isn't enough on its own. Gucci had it and still landed on the wrong side of the line. Brands that come out of this decade owning ‘human-made’ as a real asset will be the ones that treat it as a deliberate, funded choice today, not a theme to reach for once the market has already sorted itself.
Where in our production is a human hand currently invisible and should it be visible instead? Many brands already have real craftsmanship somewhere in their process and don't talk about it, because for thirty years ‘efficient’ was the flattering word. That calculation has just flipped.
If a competitor announced tomorrow that everything they do is 100% AI-generated, would our customers be relieved or alarmed? That single hypothetical sorts brands faster than any positioning workshop.
The mechanical watch didn't become valuable again because it got better at telling time. It became valuable because it stopped needing to compete on that basis at all and competed instead on being the thing a machine couldn't be, and wouldn't bother to be, because someone decided the invisible forty hours mattered. The same door is opening now, well beyond watches, and it won't stay open forever. The brands that step through it deliberately get to define what ‘human-made’ means in their category and those that wait will simply be renting the definition from someone else.
