Your Empathy Has a Maintenance Schedule

I'm not worried AI gets smarter. I'm worried we get duller.

For most of history, being smart was the moat.

That moat is closing. Faster than we are admitting to each other, and faster than most people's career plans assume. I spent years building AI products from the inside. I've watched the capability curve up close, and the honest read is this: most of what we called knowledge work is going to be done better, faster, and cheaper by something that never sleeps and never gets discouraged.

I've stopped being scared of that part. The machines getting smarter is not the thing that keeps me up.

What keeps me up is the other curve — the one nobody is charting. That while the machines get sharper, we are getting duller, and calling it keeping up.

Because if the machines take thinking — what is actually left of us?

Not a skill. Not a job title. Something older than both.

The part no model gets

The catch in your throat when a friend finally says the thing they've been carrying for months.

The way your whole body knows a room has gone wrong before anyone has spoken.

A hand on your back that puts your nervous system down like a switch.

Wanting something so badly at 2am that you get up and build it — with nobody watching, nobody asking, and no reasonable expectation that it works.

That is the territory. Feeling, presence, touch, taste, judgment under real uncertainty, the ache that makes a person make something. Almost everyone agrees this is what stays ours. It has become the standard comfort at the end of every AI panel: don't worry, machines can't feel.

And then the conversation stops right where it should start.

Because here is the part almost nobody says out loud: none of that is a mindset. All of it is biology.

Empathy is not a personality trait. It is a physiological output. Deprive a person of sleep and the connection between the emotional brain and the regulating prefrontal cortex weakens measurably. They read neutral faces as hostile. They react instead of respond. Same person. Same values. Different hardware state.

Your capacity to be moved runs on a nervous system that has been allowed to come down out of threat. Your ability to read a face runs, in part, on sleep you got two nights ago. Wanting things — actually wanting, not performing ambition on a calendar — runs on dopamine, on hormones, on a brain that got to clear its own waste overnight.

Which leads somewhere uncomfortable.

The most human parts of us are also the most perishable parts of us.

What that looks like from the inside

I've watched it happen in myself.

There was a stretch where I was under-slept, over-scheduled, and running hot for months. I was still competent. I still shipped. On paper, nothing was wrong.

But I was thinner. Less able to be moved by anything. Shorter with people I love, then quietly ashamed about it at midnight. In the room and somehow not in it. Someone would tell me something that mattered and I could feel myself processing it instead of receiving it — like there was a pane of glass where the response should have been.

That is not burnout as a productivity problem. That is the hardware for being human degrading. Quietly. On a schedule. And the cruel part is that it degrades the exact faculty you would need in order to notice.

Scale that up and you get the trade most high performers are making right now without ever naming it: burning down the biology that makes them irreplaceable, in order to keep pace with a machine they were never going to out-work.

I don't think that trade survives contact with the next five years.

Longevity, re-defined

So I've stopped thinking of longevity as adding years at the end of life.

That framing was always a little abstract and a little vain. It puts the payoff somewhere far away, for a version of you that you cannot picture and do not yet care about.

Here is the version I actually believe:

Longevity is protecting your capacity to feel the years you are already in.

That reframe changes what the work is for. Sleep is not optimization theater; it is the maintenance schedule for emotional range. Morning light is not a wellness ritual; it is how the clock that governs mood, hormones, and recovery stays anchored. Zone 2 and strength work aren't about a number on a scan; they protect the brain you will need to still be interesting at 70. Nervous system recovery isn't softness — it is the difference between reacting to your life and being present for it.

And the emerging side of this matters too. The biotech aimed at the underlying rate of biological aging — not the symptoms, the rate — is finally becoming real enough to take seriously, if you can hold both a healthy skepticism and an open mind at the same time. Lifestyle that follows the design of human biology, plus the science that extends what that design can do. Those are not competing camps. They are the same project.

The point of all of it is not a longer obituary.

It's more years of being someone who can still be reached.

What we're actually deciding

We are about to find out what humans are for. Not as a thought experiment — as a matter of ordinary economics, inside this decade.

Let the machines get better at thinking. Genuinely: let them. That race is over and it was never the interesting one.

Our job is to stay good at being human. To keep the equipment that lets us feel, connect, want, and be present in working order — for as long as biology and science together will allow.

And that was never a philosophy problem.

It's a biology problem. Which means, for the first time, it's a solvable one.