We Don’t Have an Aging Workforce. We Have an Experience Shortage: AI may be turning longevity from a demographic problem into a business strategy.
For years, the conversation has gone one way: We have an ageing workforce. The new conversation may go another way entirely: We have an experience shortage.
Same people. Same grey hair. Completely different problem. One shows up as a cost center. The other shows up as a competitive advantage nobody’s exploiting yet.
That distinction crystallized for me somewhere in Bank of Singapore’s new 2026 Supertrends: The Longevity Economy report—a document ostensibly about the business of getting old, which turns out to be a document about the business of staying useful. The bank estimates the longevity economy will grow from US$3 trillion in 2025 to US$5.4 trillion by 2034, propelled by longer lives, falling birth rates and the slow rewiring of how people work, spend and live.
You know, the same thing I have been writing about for a few years now, the thing I call The Epilogue Economy.
My Substack - Some Assembly Required - has always been where I work things out in public. It’s where the ideas are longest, sharpest, and least edited for palatability.
If you want to stay connected to where the thinking is going, not just the occasional post that survives the algorithm, please subscribe.
Most longevity reports open with the consumer pitch—the wellness retreats, the anti-ageing serums, the entire economy built around not looking your age. This one opens somewhere more interesting: the workforce itself. It argues that AI and robotics are becoming critical enablers, capable of augmenting a shrinking, greying labour pool while—in the report’s own words—“elevating the contribution of experienced talent.”
Worth pausing on who’s saying this. Bank of Singapore is not in the business of employee engagement surveys. It manages money. When an investment report starts treating experienced talent as an economic asset rather than an aging liability, I read that as more than HR sentiment. I read it as a thesis. And AI, oddly enough, may be the reason the thesis holds.
The old ledger
Walk into most workforce-planning meetings and older employees are still filed under liabilities.
They’re expensive. They sit atop the org chart, blocking the view for everyone below them. They complicate succession plans that HR has quietly been rewriting for years. Their healthcare premiums trend the wrong direction. Their skills, it’s assumed without much scrutiny, are aging out along with them. And eventually—mercifully, efficiently—they retire, making room for someone twenty years younger who costs less and, we’re assured, arrives pre-loaded with whatever the new operating system requires.
This assumption produces a very consistent set of corporate reflexes: flatten the org chart, thin the senior ranks, dangle early retirement, quietly convert institutional memory into a consulting day rate, or simply let it walk out the door—cardboard box under one arm, commemorative pen in the other, thirty years of pattern recognition gone before lunch.
Underneath all of it sits one belief:
Experience is a relic of the last operating model, not an asset for the next one.
AI is quietly putting that belief under pressure. Not because people over fifty possess some mystical wisdom unavailable to the rest of us—they don’t, and this isn’t an argument for romanticizing age. It is because AI is relocating where human value actually lives inside the work. I should know, because I am one of those people they are talking about.
The bookends have become the job
I’ve written before about how the ratio of a day’s work is flipping, and it’s worth restating because it’s the hinge this whole argument swings on.
Work used to break down roughly 10/80/10: ten percent deciding what to make, eighty percent actually making it, ten percent judging whether it was any good. The middle was the job. That’s where the headcount lived, the org charts sprawled, the budgets went.
Under AI, that ratio is drifting towards something closer to 40/20/40. Human effort is migrating to the bookends—defining the problem and briefing the machine at the front end; judging, refining and taking responsibility for what comes out the back end. The middle, the actual production, gets faster, cheaper and increasingly automatic.
Which means taste becomes the job.
So does judgement. Context. Pattern recognition. The instinct to know when an answer is technically correct and strategically idiotic—the kind of thing many people can understand in theory, but someone who has been burned by it recognises before it ships.
None of this belongs exclusively to older workers. But it accumulates through the specific, unglamorous experience of being wrong in public: watching a sound plan produce a ridiculous outcome, negotiating with someone who has no interest in a mutually beneficial deal, sensing that a client’s silence in a meeting means something isn’t landing before anyone says so out loud.
You don’t learn to read a room from a training dataset. You learn it by sitting in enough rooms.
Which makes the current moment almost comic in its self-sabotage: companies are racing to automate the middle of the work while quietly encouraging the people most practised at the bookends to take the package and go. Then everyone’s confused about why the output is fast, polished and somehow unmistakably off.
From labor shortage to judgment shortage
A labor shortage means an organization needs more hands. An experience shortage means it doesn’t have enough people who can decide what’s worth doing, supervise the systems doing it, and catch the moment a plausible-sounding answer is about to become an expensive one.
The distinction isn’t semantic. It changes the entire playbook. If the problem is labor, automation is the fix—more machine, less headcount, done. If the problem is judgment, automation doesn’t fix it. It can make it worse.
AI will generate more options than any human team could produce alone. It synthesizes, drafts, prototypes, and answers with a fluency that feels like magic—right up until one of those answers is confidently, articulately, catastrophically wrong, and nobody in the room has the scar tissue to notice before it ships.
The bottleneck hasn’t disappeared. It’s moved. The organization is no longer waiting on someone to produce the thing. It’s waiting on someone to decide which thing matters, which output can be trusted, and which small, easily missed error is quietly compounding into next quarter’s crisis.
That is where experienced people may become more valuable. Not despite the machines.
Because of them.
The real opportunity here isn’t keeping people on the payroll longer out of decency. It’s redesigning the work around the contribution they may now be unusually well positioned to make.
Experience is not tenure
Here’s the part of this argument that deserves some skepticism, including from me.
Experience doesn’t automatically convert into judgment. Some people spend thirty years sharpening their instincts. Others spend thirty years defending habits that stopped working around the time the smartphone showed up, and simply never noticed.
Unexamined experience calcifies into certainty. Seniority, handled badly, becomes a moat against curiosity rather than a source of insight. Plenty of experienced people will treat AI the way they treated the last three pieces of enterprise software: ignore it, badmouth it in the hallway, and quietly hand it off to whoever seems most comfortable with the tools.
That isn’t the experience economy. That’s nostalgia with a corner office. Bank of Singapore’s framing only holds up if experienced talent stays capable of learning alongside the systems it’s meant to be supervising. The value isn’t in the years. It’s in what the years were spent doing.
Put plainly:
Experience without curiosity is just seniority waiting to expire. AI fluency without judgment is just slop, generated faster.
Combine the two—hard-won judgment and genuine fluency with the tools—and you get something most workforce models are still not designed to recognize, develop, or reward.
Longevity as operating strategy
If experience is becoming the scarce input, longevity can’t stay parked in the DEI slide of the annual report, next to the stock photo of three generations smiling at a laptop. It has to become an operating question, asked with the same rigor as capital allocation:
Which forms of judgment are load-bearing for this business—the calls that, made badly, actually hurt?
Where does that judgment currently sit, and what is being done to keep it current and pass it on?
What happens the day the person who understands why the system works stops showing up?
Too often, organizations reduce intergenerational work to a lazy division of labor: one person operates the new tools, another supplies the war stories, and everyone calls it mentorship. That isn’t complementary capability. It is two stereotypes sitting beside the same laptop.
Companies have spent a decade and a small fortune documenting processes, building wikis, and turning tacit expertise into tidy templates. AI will make all of that easier to search and interrogate—a genuine gift.
But not everything a good operator knows lives in a document.
Some of it shows up as hesitation before a decision nobody else questioned. A raised eyebrow at a plan that looks fine on paper. A sense that two unconnected events in different parts of the business are actually the same problem wearing different clothes. The recognition that a familiar pattern is forming under conditions that only look unfamiliar.
Try writing that into a job description. Try automating it.
It’s becoming more valuable precisely because it can’t be—at the exact moment machines are fully capable of executing a bad idea beautifully, at scale, before a human finishes the sentence that would have stopped it.
The new longevity economy
The longevity economy usually gets discussed as a market created by older consumers with money and time—the retirement cruises, the second-act career coaches, the vitamin subscriptions.
That’s half the story, and arguably the less interesting half.
The other half is a productive economy, created by people living and working longer, whose experience may appreciate in value as AI changes what human contribution actually looks like.
That doesn’t happen on its own. It requires organisations willing to invest in real, continuous learning; to redesign roles instead of just relabelling them; to interrogate the age-based assumptions baked into every promotion cycle; and to build credible paths for experienced workers to actually use AI rather than merely survive its rollout with a straight face.
It requires older workers to stay curious enough to keep converting experience into judgment that’s actually current, not judgment calibrated for a market that stopped existing a decade ago. And it requires leaders to retire the fantasy that knowledge transfers itself the moment someone hands back their badge—that thirty years of pattern recognition gets automatically forwarded to whoever inherits the email account, like an out-of-office reply.
The aging workforce has always sounded like something happening to business—a bill coming due, a demographic wave to be weathered. The experience shortage sounds like something business did to itself, one early-retirement package at a time. One framing asks how much it will cost to carry people who are working and living longer. The other asks how much value is walking out the door—how much already has—because nobody thought to ask what those longer lives had actually accumulated.
One is a cost. The other, still sitting right there in the building, is an opportunity nobody’s bothered to price.
Source: Bank of Singapore, “2026 Supertrends: The Longevity Economy,” published 14 July 2026.