You Cannot AI Your Way Out of a Funding Cut

There is a version of the rural healthcare pitch going around Washington that goes something like this: small towns are losing their clinics and their doctors, so let us give them AI avatars instead. I want to be careful here, because AI genuinely can help rural care, and I will get to where it does. But the framing bothers me, because it quietly treats a structural problem, closed hospitals and vanished clinicians, as if it were a user-experience problem you can solve with a friendly screen. Those are not the same thing, and pretending they are has a cost.

The pitch: avatars for the heartland

Let me lay out the actual proposal fairly. Dr. Oz and administration officials are touting AI as a game-changing force for rural health, with tools that send nudges on prescriptions and help providers figure out which patients need care and when. On its own, none of that is objectionable. A prescription reminder is fine. A triage assist is fine. The trouble starts when an avatar is positioned as a stand-in for access itself, as if a screen in the pharmacy could replace the clinic that closed down the road. That is a big leap, and it is the leap the pitch quietly asks you to make.

You cannot AI your way out of a funding cut

Here is the number that reframes the whole conversation. Rural providers are staring down roughly 1 trillion dollars in Medicaid cuts over the coming decade. Against that, the Rural Health Transformation Program is 50 billion dollars. I am all for 50 billion dollars spent well, but put the two side by side and the AI program is about a nickel on the dollar of what is being taken away. You cannot software your way across a gap that size. When rural health leaders say the proposal addresses the symptom rather than the systemic funding hole, this is what they mean, and the arithmetic is on their side.

Where AI is genuinely weak: the subjective call

Even setting the money aside, there is a technical mismatch that worries me. AI is genuinely good at clear determinations pulled from large datasets: is a refill due, is a lab value out of range, is a risk score climbing. It is much weaker at subjective, human judgment calls. And rural coverage decisions are full of exactly those. Deciding who qualifies for a medical frailty exemption is not a clean data lookup, it is a nuanced call about a whole person. Handing that to a system that shines on the clear cases and fumbles the fuzzy ones is how you get decisions that are fast, confident, and wrong for the people who can least afford it.

The people it is for do not trust it yet

And then there is the group whose opinion should matter most: rural patients themselves. Many are openly wary of AI for personal healthcare, telling reporters things like they accept AI for some things but not their own care, or simply that they are old-fashioned and do not believe in it. That is not ignorance to be overcome, it is a signal to respect. The evidence base is thin too, only 26 peer-reviewed studies on AI in rural healthcare across fifteen years, layered on top of real infrastructure gaps: shaky broadband, thin IT staff, and clinicians with no time for one more system. You cannot push a trust-heavy technology onto people who do not trust it, over internet that does not reach them, and call it access.

How to tell help from theater

So I am not here to dunk on rural AI, I am here to give you a test. The single question that separates help from theater is this: does the AI arrive with clinicians and connectivity, or instead of them. AI plus a clinician genuinely expands what a stretched rural team can do. AI in place of a clinician just puts a friendly face on the gap. Start where it already works, like Cherry County Hospital in Nebraska using AI scribes to give overworked staff their time back. Fund the plumbing first, because no broadband means no AI. And measure the only thing that counts, whether more rural patients actually got seen. Do that, and the 50 billion dollars might do some real good. Skip it, and it is an avatar waving at an empty waiting room.