The Two Longevity Hacks Silicon Valley Can’t Sell You
Are you sleeping enough? And are you lonely?
Imagine you wanted to design a machine that would pull the maximum possible amount of money out of a person terrified of dying.
You wouldn’t build it around the interventions that work.
The interventions that work are boring, and worse than boring, they’re free, and a machine that recommends free things can’t sustain a market cap.
You’d build instead a machine that waves continuously toward the frontier of biological possibility, telomeres and senescent cells and NAD+ and rapamycin and the mitochondria of a Greenland shark, while steering the customer away from the two things that would help, because those two things generate no recurring revenue and can’t be improved by version 2.0.
For the median well-off Silicon Valley person spending money on longevity, two free interventions dominate the entire effect of everything they buy by a wide and somewhat embarrassing margin:
Get enough sleep, and…
Don’t be lonely.
Everything downstream of those two things is, at best, rounding error, and at worst, a way to feel like you’re doing something so you don’t have to do the hard things.
Sleep duration and all-cause mortality trace out a U-curve: too little kills you, and although too much tracks with dying too, the “too much” arm probably reflects sick people sleeping more rather than sleep causing the sickness. The bottom of the U falls around seven hours. People who habitually sleep five hours or fewer show elevated mortality risk in cohort after cohort.
When you deprive yourself of sleep, you degrade essentially every biomarker the companies sell you interventions to improve. Sleep too little for long enough, and you raise your inflammatory markers, worsen your insulin sensitivity, elevate your blood pressure, and impair your glucose tolerance to the point where a week of restriction can make a healthy young man’s metabolic panel look pre-diabetic.
During deep sleep, your brain runs its own cleanup crew, the glymphatic system, which flushes out metabolic waste including the amyloid-beta everyone frets about.
Now take the guy who takes berberine to improve his insulin sensitivity, swallows an anti-inflammatory nutraceutical stack to lower his inflammatory load, wears a device that buzzes when his heart rate variability drops, and spends $400 a month on the whole apparatus, and then sleeps six hours a night because he’s optimizing. He’s bailing water out of a boat with a very expensive artisanal bucket while declining to plug the hole, which would have cost nothing, because plugging the hole would just mean going to bed before ten.
Weak social connection carries a mortality risk comparable to smoking, and larger than obesity or physical inactivity. Across hundreds of thousands of people, the survival benefit of social connection roughly matches that of not smoking 15 cigarettes a day.
I’ll admit social connection is horribly confounded. Sick people withdraw from social life; poor people have thinner social networks and worse everything else; the personality traits that predict isolation also predict a hundred other things. Reverse causation and hidden confounders could be doing a great deal of work. No randomized trial proves that acquiring friends adds X years to your life, and none ever will, because you can’t randomize people into loneliness for forty years and count the corpses, thank god. But across an enormous range of study designs, populations, and attempts to control for the obvious confounders, the association holds up better than almost anything else in epidemiology; it survives adjustment where most findings crumble, and it runs through mechanisms nobody disputes. People who stay lonely for years have higher cortisol levels, more inflammation, sleep worse, and put their bodies through everything chronic stress does.
The evidence that friends and sleep extend your life beats the evidence for almost anything you can buy: larger effect sizes, more replications, longer follow-ups, more biological plausibility. Rapamycin has real, striking data in mice and essentially zero long-term mortality data in healthy humans. Metformin’s much-hyped longevity signal comes largely from an observational comparison where diabetics on metformin appeared to outlive non-diabetics, which should make you deeply suspicious rather than deeply excited, and TAME, the randomized trial meant to test it, has struggled for years to find funding. NMN and NR raise NAD+ in your blood, which sellers wave around as if it were the goal rather than a stand-in for it, and the human healthspan data remain thin. Some of these compounds might work, and reasonable people find rapamycin interesting enough to run real trials. But the companies market these things with a confidence that runs inversely to the strength of the evidence, and the two free interventions with the strongest evidence draw the least attention, and that lines up too neatly to be an accident.
Why isn’t it an accident?
I don’t think the supplement companies are lying to you, though I’m sure some of them are.
But I think they, and the various influencers, have created entire product lines and movements that soak up the fear that would otherwise force you to confront the boring interventions.
You fear death, and because you fear it you demand action, the sensation of Doing Something About It. Sleep and friendship do a terrible job of supplying that sensation, because they don’t feel like doing. Going to bed means stopping, and calling your friend means doing something you already knew you should do and had felt vaguely guilty about skipping. There’s no package, protocol, dashboard, or number that climbs.
The supplement gives you a ritual to follow instead. You take it at the same time each day, you read forums about it, you stack it with other things, and it gives the fear something to grip onto.
The companies sell you the sensation of having handled the problem, and that sensation shoves aside the interventions that would actually work, because once you’re already Doing Something you feel less pressure to do the hard, boring, unpackageable ~other somethings.
If the supplements were merely useless, they’d tax the anxious, and they’d be annoying but harmless. But you have only so much fear to spend; you can sustain only so much active dread about your own mortality before you have to discharge it, and once you discharge it onto a pill it’s gone, no longer available to power the harder projects of rebuilding your evenings so you sleep enough and see the people you love.
Knowledge isn’t the bottleneck; we already know all of this. Nobody spending $400 a month on a longevity stack has failed to hear that sleep matters or that loneliness corrodes you. But the free interventions are hard; you have to accept that you’ll be less productive in the short term, you’ll have to make phone calls that feel awkward, and you'll have to build a life around unglamorous constraints. The paid interventions are easy: they only require a credit card and a shelf.
Before you spend another dollar optimizing the margins, check whether you’ve maxed out the two interventions that dominate the whole field.
Do you sleep seven-ish hours, reliably, or do you sacrifice sleep to the very productivity that supposedly funds your quest to live forever?
Do you have people, not followers, not colleagues, actual people, whom you see and talk to often enough that a crude epidemiological instrument would register it?
If either answer is no, then everything else you’re doing amounts, statistically, to very sophisticated procrastination. If both answers are yes, if you’ve genuinely got the sleep and the friendship handled and you still want to take rapamycin and get your NAD+ measured, then go ahead.
Just don’t mistake the frontier for the foundation.