I spent two years building a company's clinical infrastructure, and I never had a title.
Nobody was withholding one. There simply wasn't a word for what I did.
I wrote the training. Built the workflows. Assembled the regulatory libraries. Designed the forms the clinicians actually touched. On any given day, I could move from a clinical question to a software decision to whatever operational problem had appeared that morning.
Five departments would have had to share me.
So none of them did.
When it came time to renew, the company offered me equity.
I asked for cash instead, along with a few terms I had already raised six months earlier. The answer was that nothing in the document could change.
Investors.
So I left.
I've been thinking about that conversation for three weeks. At first, I thought it was about money.
Now I wonder if equity was simply the closest thing they had to a price for work they couldn't classify.
There was no row for the work.
So where would the rate go?
Most companies are built around specialization.
First you become a doctor. Then a cardiologist. Then an electrophysiologist. Eventually, you become the person everyone calls when a very specific electrical signal in a very specific part of the heart starts doing something weird.
Medicine became extraordinarily good by dividing the body into systems, specialties, and increasingly narrow areas of expertise.
We organized companies the same way.
Clinical goes here. Product goes there. Operations gets another department. Every role has a title, a manager, a budget, and a list of responsibilities that is supposed to fit inside one box.
That structure makes sense.
For most of modern working life, one person could not know enough to move easily between fields. Knowledge lived in universities, journals, laboratories, apprenticeships, and inside the heads of people who had spent decades collecting it.
If you needed something beyond your expertise, you needed another expert.
AI has made it much easier to reach into an adjacent discipline.
It can help you understand a method you have forgotten, summarize a field you don't work in, or write the code you would never attempt alone. You can learn enough to ask a better question and recognize who needs to be in the room.
You still need the expert.
You also need someone who can see how that expertise connects to everything around it.
Think about the human body.
Your heart is highly specialized. So are your kidneys and your liver. Each one performs work the others cannot replace.
Something still has to connect them.
Something has to notice what changed over here and understand why it matters over there.
Companies are very good at hiring the organs.
They have a harder time hiring the nervous system.
There is one place where this kind of person fits naturally: an early-stage company.
When eight people are building something together, everyone does six jobs. The person who can move between a clinical question, a product decision, and an operational mess in the same afternoon is often the most useful person in the room.
Nobody pauses to call them a generalist.
They are simply doing what needs to be done.
Then the company raises money. It hires more people. Departments appear. Job descriptions get narrower. Work becomes easier to assign and measure.
Eventually, the person who crossed the whole company is asked to choose a lane.
The person who made sense at eight people becomes difficult to place at eighty.
I don't think that's a betrayal. The company has entered a different phase and needs more structure.
But I do think it explains where some of these people go.
A surprising number of them end up working for themselves and calling themselves creators.
I've never loved the word.
Many of the people I see using it do much more than make content. They write, advise, teach, build products, solve operational problems, and connect ideas from several industries.
Their work crossed too many lanes for one job description.
So they built their own structure and gave themselves a title broad enough to hold it.
It's especially interesting now because companies are trying to figure out what to do with AI.
The technology is moving quickly. Organizations move through meetings, budgets, approvals, policies, and departments that don't always agree on who owns the decision.
In healthcare, an AI project can sit somewhere between clinical, technical, operational, legal, and commercial teams.
Everyone owns a piece.
The full problem lives in the space between them.
A model may be able to perform the task. Someone still has to decide whether the task matters, where it belongs in the workflow, what could go wrong, who carries the risk, and whether the person expected to use it on Tuesday morning will actually use it.
That work requires more than access to information.
It requires judgment.
Judgment comes from watching decisions become consequences. From building the form, watching someone use it incorrectly, rewriting it, and eventually realizing the original problem was never the form.
AI can help people move between disciplines faster.
It cannot give them the experience that tells them which answer belongs in the room.
Within five years, companies building clinical AI will need more people who can hold clinical, technical, operational, and commercial context at the same time.
I don't expect them to create real roles for it.
They'll keep hiring that capability from outside, under whatever title is available: strategist, advisor, consultant, implementation lead, writer, fractional executive.
If those roles do get built, with budget and authority and a name, then the org chart adapted faster than I think it can, and I'm wrong.
I'd be glad to be.
Until then, people who cross disciplines will keep looking difficult to place inside companies and unusually useful outside them.
Which makes the missing row worth paying attention to.
Sometimes your work becomes valuable before an organization develops the language to describe it.
So here is the question I'm sitting with:
What did you stop being curious about because someone told you it wasn't your lane?
Go pick it back up.
Did you enjoy this email?
|
