It is one of the quiet scandals of modern science that we still have no generally accepted theory of aging. We can name the exact pathogen behind an infection and the exact thermonuclear reactions warming the cores of distant stars, but ask a biologist what, precisely, aging is, and the room divides into a dozen camps, each with its own pet mechanism, its own molecule, and quite often its own startup. There should be a theory of aging. There is none. Our knowledge here looks less like a map with a few blank corners and more like a black hole.

I want to suggest that the trouble lies in the question itself. We have been studying the shadow and calling it the object — the hole, and calling it the donut. Suppose aging, as a single thing, does not exist at all. What exists instead is a crowd of resilience mechanisms: DNA repair, immunity, apoptosis, the renewal of tissue from stem cells, the stubborn mechanical integrity of the cellular matrix. “Aging” is simply the name we give to the moment when they all give way at roughly the same time — and that simultaneity, I will argue, is no accident. It is engineered, slowly and blindly, by natural selection.

The factory and the warranty period

Imagine a factory building a car designed to last 100,000 kilometers. A sensible engineer does not buy a transmission rated for 300,000 km and bolt it to brake pads that fail at 40,000. The expensive over-built part is wasted: the car will be in a scrapyard long before that transmission shows its quality. Equally, no part should be rated below the target, or the car dies young and the warranty bleeds money. So the rational move is to source every component with roughly the same service life — a hair above the design target, and no more.

Evolution is exactly this kind of frugal, indifferent purchasing department. If any one resilience system fails earlier than the lifespan an organism needs to reproduce, selection works furiously to extend it. And any system that lasts conspicuously longer than needed is left undefended: harmful mutations accumulate in it unopposed, because no individual ever lives long enough for those mutations to matter. This is Medawar’s old “shadow of selection,” but pointed not at single genes — at whole stability systems. Over evolutionary time the durable mechanisms drift down and the fragile ones are pushed up, until they converge. Every system in the body is issued the same warranty period.

Why you cannot cut off just one head

Picture an animal that, left alone, would die of atherosclerosis at 50, of cancer at 70, and of Alzheimer’s at 80. Now drop it into a niche that rewards longer life. Selection attacks atherosclerosis first, because almost nobody currently survives to meet the cancer. Meanwhile nothing whatsoever selects against Alzheimer’s — no one lives to see it, just as cave fish lose their eyes for lack of any pressure to keep them. Run this forward and you get an animal in which atherosclerosis, cancer, and Alzheimer’s all arrive at around 70. The instant one cause tries to dominate, selection planes it back down, like a carpenter’s blade meeting a knot, and levels it with the rest.

This is why, in humans, unrelated families of age-associated disease — heart, brain, malignancy — cluster around the same decades and carry roughly equal weight, each claiming its 20 to 30 percent. No single system was built to last 300 years; they all wear out together. The salmon shows the boundary case: selection never cared how it would die, only that it die promptly after spawning so the young could feed, so its repair systems are simply switched off and it rots while still alive. The naked mole rat shows the other extreme — selection so relentless that it stacked on extra defenses (enhanced cancer protection, a peculiar hyaluronic acid, a slowed metabolism) and pushed senescence toward zero. The elephant, carrying around twenty copies of the tumor-suppressing P53 gene where we carry two, is the same story written in a single gene.

The illusion of a master clock

Here is the trap the synchronization sets for us. When dozens of mechanisms expire in concert, it looks as though a single hidden conductor is calling time — some master clock we need only find and switch off. So we go looking for it. We look for antagonistic pleiotropy and find it. We look for an aging program and it, too, seems to work a little. We measure methylation, telomeres, hormones, and each correlates beautifully with age. But in an organism where everything decays, everything correlates with everything, and correlation tells us almost nothing about a single cause — if a single cause exists at all. The very multiplicity of resilience mechanisms generates the multiplicity of mutually contradictory “theories of aging.” Each theorist has caught a real thread and mistaken it for the whole rope.

The hard consequence is blunt: no single mechanism means no single cure. Try to kill aging by attacking one mechanism and you meet the hydra — cut off one head and seven grow. This is also why interventions that add tens of percent to a mouse’s life do so little for us. Humans were under heavy selection for longevity for millions of years — we roughly doubled the chimpanzee’s lifespan to make room for grandmothers and for old sages who carried a culture’s memory before writing existed. All the easy levers have already been pulled by evolution. The low-hanging fruit was harvested long ago.

So what do we actually do?

If despair were the only response, this would be a short and gloomy essay. But the absence of an essence can be turned from a curse into a recipe. If aging is many things, fight it with many things at once.

First, bolster every major resilience system simultaneously. There is a principle the rationalist community calls “More Dakka” — if a little of something helps, try a lot more of it. Chemotherapy failed until doctors combined drugs and pushed the dose; HIV stayed lethal on one antiviral and became survivable on three. One drug: death. Three or more: life. The number of stability mechanisms is large but finite — perhaps ten to a hundred. Find ten or a hundred safe geroprotectors, each striking a different pathway — one on the matrix, one on insulin-like growth factor, one on the epigenetic clock — and aging may suddenly slow in a way no single molecule ever could.

Second, hunt down the accelerators. Smoking, metabolic syndrome, hypertension, chronic inflammation, sedentary internet-bound days — these are not the engine of aging, but they load the car so it crawls. Remove the dead weight and you buy time cheaply.

Third, lean on what looks like second-order repair — repairing the repair systems themselves. Nature built almost none of this, which is precisely why it is such fertile ground for us. Sleep behaves like a whole-body maintenance crew. Stem-cell renewal, when we learn to control rather than merely inject it, is repair of the repairers. The state has its internal-affairs bureau to police its police; the body, mostly, does not — so we will have to supply one.

And finally, the long horizon. True repair demands a tool as complex as the thing it mends: to mend a soldering iron you need another soldering iron. The ideal anti-aging device would be a swarm of bionanorobots directed by AI, capable of diagnosing and fixing damage faster than it accumulates — or, failing that, the replacement of failing parts outright, up to and including the body. We are not there. But we may not need to be there yet.

The warranty is a schedule, not a wall

The deepest point is this: the warranty period is not a wall. It is a schedule — and schedules can be renegotiated. The probability that civilization develops genuine life-extension technology is not spread evenly across the years ahead; it grows, steeply. So even a modest deceleration of aging is worth far more than it looks, because it raises the odds of being alive on the day the real breakthroughs arrive. A few extra percent of lifespan can be the difference between just missing the boat and reaching longevity escape velocity — the point at which each year of research buys more than a year of remaining life.

The most important question was never whether aging can be defeated in the abstract. It is whether it can be defeated within our own lifetimes. And the honest answer is that we do not have to find the donut. We only have to keep enough of its many vanishing edges intact, for long enough, to live to see the bakery rebuilt.

Adapted from the essay “No Theory for Old Man: Evolution Led to an Equal Contribution of Various Aging Mechanisms” by Alexey Turchin (Open Longevity, 2023).