The Shift No One in Healthcare Wants to Talk About

Over the past few weeks, I have been hearing a lot of messaging around AI that honestly does not sit right with me.

A lot of reassurance. A lot of calming language. A lot of: "Do not worry. AI is not coming for your job."

And the more I hear it, the less I believe it.

The Narrative Shift

Not long ago, the tone was completely different. AI companies—and the people backing them—were openly saying this would change everything. Entire industries. Entire roles. Entire workflows.

And then, almost overnight, that messaging shifted. A lot.

Now it is: "This is just a tool." "This will help, not replace." "Nothing to worry about."

I do not know if that shift is coming from optimism, ignorance, or something a little more coordinated. But either way, I think it is wrong.

And if you work in healthcare, you can already feel why.

This Is Not Some Distant Future

It is already happening—quietly, but consistently—in the background of everyday workflows:

• Documentation is being automated
• Scheduling is being automated
• Intake is being automated
• Triage is being automated
• Administrative layers are being compressed

None of this is theoretical. It is here. And it is accelerating.

This is not about fear. It is about clarity.

The Questions We Are Still Asking

Because at the exact same time all of this is happening, we are still asking the same questions:

• Why are providers more burned out than ever?
• Why does access to care still feel broken?
• Why does it feel harder—not easier—to actually take care of patients?

The Relationship at the Core

At its core, healthcare has always been simple: A provider. A patient. A relationship.

That relationship is sacred. Everything else in healthcare is supposed to support it.

But over time, we have built layer after layer around that relationship: administrative systems, insurance complexity, documentation requirements, fragmented referral networks, technology that often adds more steps instead of removing them.

Most of these layers were built with good intentions. But today, they often stand between providers and the people they are trying to care for.

AI Enters the System

And now AI is entering that system.

On one hand, it has the potential to remove a lot of that friction: less time documenting, less time navigating systems, more time actually caring for patients. That is the promise.

But there is another side to this that people seem much less willing to talk about.

AI does not just make systems more efficient. It changes where value exists inside those systems. It compresses certain roles. It reshapes workflows. It shifts how organizations think about staffing.

And healthcare is not going to be immune to that.

The Real Question

We do not need to panic. But we do need to be honest: the structure of the healthcare workforce is going to change—and it is going to change quickly.

The real question is not: "Is AI replacing healthcare jobs?" That framing is too simple.

The real question is: "Which parts of healthcare should never be automated—and where are humans needed most?"

Where Humans Are Needed Most

Because when you zoom out, there are entire areas of care that are not shrinking—they are growing. And they are deeply, irreducibly human:

• In-home and longitudinal care
• Chronic disease management
• Behavioral and mental health
• Community-based and underserved care
• Patient navigation and care coordination

These are not edge cases. These are some of the biggest gaps in our system today. And they all require the same thing: trust, judgment, human connection.

The Path Forward

This is where I think the conversation needs to go next. Not fear. Not denial. But clarity.

Because there is a real opportunity here. For the people building in this space—whether in AI, healthcare systems, or public health—there is a responsibility to get this right.

To help healthcare workers adapt to these changes. To create pathways into roles where human care is actually needed. To ensure that efficiency gains do not just reduce costs, but improve care. To reinvest time and resources back into the provider–patient relationship.

If this transition is handled thoughtfully, it could strengthen the healthcare workforce. If it is handled poorly, it could do the opposite.

An Initiative at the Intersection

That is why I have started working on an MPH-led initiative focused on this exact intersection: where access to care is actually breaking down, how providers are experiencing these changes in real time, and where AI can help—without replacing what matters most.

This work connects directly to projects like QClinics Solutions, which is building infrastructure to strengthen clinic operations and improve care delivery in underserved communities.

Because the goal is not just to build new tools. It is to make sure that whatever comes next still centers around the relationship that healthcare was built on in the first place.