The Highest Use of Us

Prefer to listen in your podcast app? Apple Podcasts · Spotify

There is a particular feeling that comes from doing something you are exceptionally well suited to do.

You are teaching and suddenly know exactly how to help someone understand. You are caring for a patient and notice the small thing that changes what happens next. You are building something and three ideas that had been floating separately suddenly connect. You are leading a team through a problem without a clear answer and find the question that opens the way through.

Sometimes we call it flow. Sometimes creativity. Sometimes purpose.

Whatever we call it, there is something deeply human about the experience of being fully used.

Not used up.

Fully used.

Your judgment. Your imagination. Your accumulated experience. Your capacity to connect things that have never been connected before. Your ability to teach, reassure, challenge, invent, notice, care.

This is where dignity lives.

And that is important to me as I think about AI in healthcare, because we are spending an enormous amount of time asking where AI belongs.

Should it document the visit? Review the chart? Answer messages? Identify risk? Support diagnosis? Manage operations? Communicate with patients?

Those are necessary questions.

But I wonder if we’re starting one question too late.

Before we decide where AI belongs, perhaps we need to decide why we are bringing it in at all.

My answer is increasingly this:

To maximize human potential.

Not productivity. Not fewer people. Not even efficiency, although efficiency may be an outcome.

The purpose is to build infrastructure capable of carrying the work that does not require the highest expression of a human being, so human beings have more capacity for the work that does.

That distinction matters because there is a difference between human work and work that is currently being done by humans.

Human work is judgment. Discernment. Reassurance. Teaching. Noticing. Relationship. Creativity. The willingness to carry moral responsibility for a decision that doesn’t have a clean answer.

Work currently being done by humans is remembering a deadline. Finding a document. Re-entering the same information a second or third time. Reconciling a spreadsheet against a system that doesn’t talk to another system. Holding a mental map of five disconnected systems so nothing between them gets lost.

Out of necessity we have conflated them, so that we sometimes treat removing the second category as though we are diminishing the first.

The opposite may be true.

When we build systems that absorb administrative burden, we create more room for the best expression of our uniquely human qualities.

And that makes me think about Frances Perkins.

March 25, 1911. Washington Square, New York City.

Perkins was across the park when the alarm bells started. She ran toward the sound and stopped at the edge of a crowd, looking up at flames pouring from the eighth and ninth floors of the Triangle Shirtwaist Factory.

There was no fire escape built to hold that many people. The stairwell doors had been locked, standard practice at the time, to keep workers from slipping out early or walking off with scraps of fabric. So the young women trapped on the upper floors had two horrific choices: the fire, or the window.

Close to fifty of them chose the window.

By the time it was over, 146 workers were dead, most of them young immigrant women.

Nothing about that day called for more heroism from the workers. No amount of individual grit would have unlocked that door.

The problem wasn’t that the people inside the system had failed to perform at a high enough level. The system had assigned them each responsibility for something no individual human being should have been responsible for carrying.

Perkins spent the next twenty-four years changing that.

She helped build laws governing fire safety, building codes, and working hours. Later, as the first woman to serve in a U.S. Cabinet, she became a principal architect of the Social Security Act.

She built structures that carried things individuals had previously been expected to somehow manage for themselves.

And here is the distinction that matters now:

Those structures didn’t make people less capable. They created conditions in which people could direct more of their capability toward the things that actually required them.

That is the opportunity I see in front of us again.

This time, the infrastructure we are deciding how to build is intelligent.

Picture a supervisor today.

Not a struggling one, an excellent one.

She knows how to develop people. She notices when someone is drowning before they say so. She can tell the difference between a mistake that needs a conversation and one that needs a policy. Given the space, she would spend her day doing exactly that.

Instead, she spends part of it tracking who is due for a certification renewal. Chasing a form a vendor portal lost. Reconciling a spreadsheet against a system that doesn’t talk to another system. Recreating evidence of something she already knows happened because the software that should have recorded it didn’t. Checking, and checking again, that nothing has fallen through a gap that exists only because no one built a floor there.

She is not overwhelmed because she is weak.

She is overwhelmed because she is the floor.

But overwhelm may not even be the most important loss. Look at what we are not getting from her while she is down there holding up the floor.

What isn’t she noticing? Who isn’t she mentoring? What problem isn’t she solving? What idea never gets enough uninterrupted attention to become an idea at all?

That is the cost we underestimate when we talk about administrative burden. We measure the hours lost to the mechanics. We rarely measure the human capacity that never had the room to emerge.

There’s an old engineering idea called mistake-proofing. The premise is almost embarrassingly simple: human beings will make errors, so instead of demanding endless vigilance, you design the process so some errors become harder, or impossible, to make.

A microwave won’t start while the door is open. It doesn’t need you to remember. The system remembers for you.

That isn’t about removing people from the work. It’s about refusing to spend human attention on a job the system can do so that attention remains available for the work only a person can do: noticing what no sensor would catch, solving a problem no one has seen before, making a judgment when the answer isn’t obvious.

Healthcare isn’t an engineered process, at least not entirely. Nor should it be.

But we make an engineering decision every time we decide what a person has to hold in her head and what the system will hold for her.

And in a lot of healthcare organizations, the answer is still: the person holds almost all of it.

We have called that dedication. Maybe some of it is just infrastructure we never built.

Good infrastructure doesn’t make responsibility smaller. It makes responsibility more human.

The system can remember, connect, prompt, document, detect, and coordinate.

The person can interpret, decide, teach, comfort, challenge, create, and lead.

That’s not a demotion. It is a chance to spend more of our working lives closer to the edge of what we are capable of.

For most of my career, that kind of infrastructure was a nice idea and a hard engineering problem. Connecting what happens across all the disconnected systems a clinic runs on required either an army of people doing it by hand or software too rigid to survive contact with how care actually happens.

Something has changed.

And I don’t think the most interesting thing about AI is how impressive the technology has become.

The interesting thing is that we now have infrastructure capable of taking on more of what has been consuming human attention because there was nowhere else to put it.

That isn’t really a claim about what AI can do in healthcare. It’s an answer to why it belongs there at all.

Not because it is efficient. Because it can create the conditions for more human potential to be expressed. And I wonder what all of the humans in healthcare will create when we are free from the administrative burdens that tie us to the system as we know it.

Infrastructure should carry the non-negotiables so leaders can carry the humans.

We describe a good leader as someone who gives feedback, develops people, catches problems early, builds accountability, and helps an organization learn.

But a leader can only do the human part of that job with whatever capacity is left after she finishes administering the mechanics of it.

Give the mechanics to infrastructure and you don’t just save her time. You give her back to the humans she leads.

And perhaps you give something else back too: the space to think beyond the immediate task. To connect ideas. To teach. To invent. To become better at the work instead of keeping up with it.

The same is true in the exam room.

If clinicians and staff are spending real cognitive and emotional capacity compensating for operational fragmentation, that cost doesn’t evaporate. It goes somewhere.

Some of what patients experience as rushed or inconsistent care, or a clinician who seems distracted might not begin in the exam room at all.

It might begin several steps upstream, in a system that never built anywhere else for five things to live except a capable person’s head, leaving less of her available for the sixth thing:

The person in front of her.

And when we give some of that capacity back, the opportunity is larger than a calmer visit.

We create room for clinicians to notice, teach, connect, wonder, and think. For administrators to build rather than patch. For leaders to develop people rather than chase evidence that the development happened.

Those are not soft benefits around the edges of healthcare. They are where learning happens. Where better care gets imagined. Where organizations evolve.

Here is the idea I keep returning to:

Maybe dignity is not only measurable by what a system does not require a person to carry. Maybe it is also visible in what that person becomes capable of doing when the unnecessary weight is gone.

The question of where we bring AI into healthcare has to be addressed. But applying the question of why we bring AI into healthcare gives us context and purpose for what we decide to do.

It isn’t really about software.

It’s about what we think a person is for, and whether we are building these new systems to protect and expand the fullest expression of that, or spend it down.

The workers who came after Perkins’s reforms didn’t have to be braver than the ones who died in 1911. They didn’t have to become better at reading buildings for danger.

They got to work inside a floor someone else had already built.

Codes that made the doors open. Inspections that made the fire escapes real.

The system carried what no individual should have had to carry alone.

And that left the person to carry what was actually hers: the skill, the judgment, the work, the life she went home to at the end of a shift.

We have another chance now to decide what belongs in the infrastructure and what we want to preserve for people.

If we get that decision right, the measure of good AI in healthcare will not be how much human work it replaces.

It will be how much human potential it releases.

So here’s the question I’d leave you with:

How much of what you carried today actually needed you? And what might become possible if more of you were available for the work that did?

Seen this pattern where you work? Hit reply and tell me. That's the harvest.

Want a bit more? Listen on Apple Podcasts or Spotify for the stories and One More Signal about this topic.