The Oracle of Time

There was a consultant physician on our medical rotation who everyone thought was behind the times. Dr Nair had been practising for over thirty years. While the rest of the department rushed to adopt whatever the latest NICE guideline recommended, he would sit back in the ward round, arms folded, and ask the same question every time: "Show me the ten-year data."

It drove the registrars mad. They would cite the newest randomised controlled trial, the freshest meta-analysis, and he would nod politely and say, "Interesting. Let's see where it is in a decade." As a junior doctor, I found it frustrating. It felt like stubbornness dressed up as wisdom. The evidence was right there. Why wouldn't he act on it?

But then I started paying attention to what happened next. The aggressive blood sugar targets in type 2 diabetes that were supposed to reduce cardiovascular events? The ACCORD trial showed they increased mortality.^[1] The blanket prescription of hormone replacement therapy to prevent heart disease in postmenopausal women? The Women's Health Initiative reversed that consensus almost overnight.^[2] The dietary fat guidelines that told an entire generation to replace butter with margarine? Quietly walked back after decades of dubious evidence.^[3]

Meanwhile, the treatments Dr Nair trusted most were the oldest ones. Rest. Fluids. Time. Early mobilisation. Reassurance. He would discharge patients with instructions that a Victorian physician would have recognised. And his patients did well. Remarkably well. He wasn't ignoring evidence. He was applying a filter that the rest of us hadn't yet learned to use. Time itself was his most trusted diagnostic tool.

I didn't have a name for what he was doing until I read Taleb. Dr Nair was practising the Lindy Effect, and he was doing it in a white coat.

The Lindy Effect

The Lindy Effect is named after Lindy's delicatessen in New York, where comedians supposedly observed that the longer a Broadway show had been running, the longer you could expect it to continue. Taleb formalised this into a broader principle: for non-perishable things, life expectancy increases with every additional day of survival.^[4]

This is counter-intuitive because it reverses everything we know about living things. A seventy-year-old human is closer to death than a twenty-year-old. But a seventy-year-old technology is further from obsolescence than a two-year-old technology. A book that has been in print for a century will almost certainly outlast one published last month.

The reason is survival itself. Every year a book stays in print, an idea stays relevant, or a technology remains in use, it has passed another round of testing by the harshest judge there is: reality. The things that survive are not random survivors. They have earned their longevity by being genuinely useful, genuinely true, or genuinely adapted to human needs.

Why the new is fragile

New things are untested. This is not a moral judgement. It is a statistical fact. A drug that has been on the market for two years has been exposed to a tiny fraction of the patient population, a narrow range of interactions, and a limited window for long-term effects to manifest. A drug that has been in use for sixty years has been tested by millions of bodies across decades of real-world conditions.

Consider Vioxx. When rofecoxib was approved, it was celebrated as a breakthrough in pain management. Selective COX-2 inhibition. Fewer gastrointestinal side effects. The clinical trials looked excellent. It was prescribed to over 80 million people worldwide before it was withdrawn in 2004, after evidence emerged that it doubled the risk of heart attacks and strokes.^[5] Aspirin, which Vioxx was designed to improve upon, has been in clinical use since 1899. It is still here.

The pattern repeats across medicine. Thalidomide. Baycol. The initial enthusiasm for routine episiotomy, for tonsillectomy in every child with recurrent sore throats, for bed rest after heart attacks. Each of these was the "latest evidence" at some point. Each was revised or reversed by time.

The newest study is precisely the one most likely to be overturned. The newest technology is precisely the one most likely to be replaced. This is not cynicism. It is the Lindy Effect working in reverse: things that are new have not yet passed the filter of time, and most of them will not.

The ancient is robust

If the new is fragile, the ancient is robust. Not because old things are inherently good, but because time has already removed the harmful variants through via negativa.

Walking. Fasting. Sleep. Clean water. Social connection. Sunlight in moderate doses. These "interventions" have survived thousands of years of testing across every human culture. They have been filtered by the most relentless process we know: natural selection acting on behaviour across countless generations. The practices that harmed their practitioners were abandoned. What remains has earned its survival.

Healthcare is addicted to the new. Every conference promises a breakthrough. Every startup pitches a revolution. But when you look at the evidence dispassionately, the most effective interventions for the majority of chronic disease are profoundly old. Exercise reduces all-cause mortality more reliably than most pharmaceuticals.^[6] Sleep deprivation is a risk factor for nearly every major disease category. Social isolation carries mortality risks comparable to smoking.^[7]

None of these require a prescription, a device, or a subscription. They require only the discipline to do what humans have always done. The Lindy Effect suggests we should trust them precisely because they are ancient.

Lindy as a filter

The Lindy Effect is not a law of physics. It is a heuristic: a rule of thumb for decision-making under uncertainty. And like all good heuristics, it is most valuable precisely when you have the least information.

When evaluating something new, ask a simple question: how long have the alternatives survived? If you are considering a new drug, ask how long the existing treatment has been in use. If you are adopting a new management framework, ask how long the one it replaces has existed. The burden of proof should always be on the new, not the old.

This does not mean rejecting innovation. It means being appropriately sceptical of novelty and appropriately respectful of longevity. A new drug that genuinely improves on a sixty-year-old treatment may be worth adopting, but the evidence threshold should be high, because you are betting against something that has already passed six decades of testing by reality.

The practical application is straightforward. When someone tells you that something is "outdated," ask: outdated by what? If it has been replaced by something with a longer track record, fair enough. If it has been replaced by something that arrived last year, be cautious. Time has not yet had its say.

The Lindy test for healthcare innovation

This is where the Lindy Effect becomes directly relevant to those of us building in health technology.

Most health tech startups are building something that did not exist five years ago to replace something that has existed for decades. The electronic health record replaces the paper chart. The AI diagnostic replaces the experienced clinician's pattern recognition. The remote monitoring platform replaces the ward round.

The Lindy Effect does not say these innovations are wrong. It says: be careful. The thing you are replacing has survived for a reason. Your innovation has not been tested by time. Understand what you are displacing and why it endured before you assume you can do better.

This does not mean we should not innovate. It means we should innovate with humility. The paper chart survived because it is simple, reliable, and requires no electricity. The ward round survived because physical presence reveals things that cameras and sensors miss. The experienced clinician's intuition survived because it encodes decades of pattern recognition that no algorithm has yet replicated at scale.

When we build technology that augments these Lindy-tested practices rather than replacing them, we are working with time rather than against it. When we build technology that eliminates them entirely, we are making a bet that our two-year-old invention is smarter than centuries of accumulated adaptation. Sometimes that bet is correct. More often than we admit, it is not.

What Lindy means for how we learn

The Lindy Effect applies not just to technologies and treatments but to knowledge itself. Taleb's own reading is heavily weighted toward texts that have survived centuries. The Stoics. Seneca. The pre-Socratics. Religious texts that have endured millennia. His argument is simple: if a book has survived five hundred years of readers, it contains something genuinely valuable. If it was published last month, you have no such assurance.

This has changed how I read. I still read new publications and follow current research. But I weight my attention differently now. When I want to understand leadership, I read Marcus Aurelius before I read the latest business bestseller. When I want to understand human nature, I read the Mahabharata before I read a psychology paper. The old texts have survived because they describe something true about the human condition, something that has not changed in two thousand years and is unlikely to change in the next two thousand.

The best education is not the most current. It is the most enduring.

The coherent framework

This is the final essay in this series, and it is worth stepping back to see how the six ideas fit together. They are not separate concepts. They are facets of the same worldview.

Antifragility teaches us to build systems that gain from stress. Fat tails warn us that extreme events are more common than we think. Skin in the game insists that decision-makers bear the consequences of their decisions. Via negativa shows us that improvement often comes from removing harm rather than adding intervention. Optionality reveals that the best strategies do not require prediction; they require asymmetric positioning.

And the Lindy Effect ties them all together with a single principle: respect what has survived. Time is the filter that catches what all our analysis misses. The treatments, ideas, institutions, and practices that have endured centuries of human use have passed a test that no clinical trial, no peer review, and no expert panel can replicate.

Dr Nair understood this. He did not have the vocabulary of Taleb. He did not need it. He had something better: thirty years of watching the new arrive with fanfare and depart in silence, while the old, the tested, the enduring, quietly continued to work.

"Show me the ten-year data." It is, I now realise, the most Lindy question a doctor can ask.

Related reading:

• The Systems Paradigm - Understanding how enduring principles shape system design
• Why Less is More - The principle of removal as a path to quality
• Incerto: The Discipline of Less - Via negativa and the power of subtraction
• Incerto: Antifragile by Design - Why systems should gain from stress, not merely survive it