We need to start talking about overhauling healthcare and education with AI

Doctors and teachers are desperate for help. This won't be a quick fix or an easy lift, but it's critical that we start work ASAP.

From Sam Altman tweeting that “doctors tell me they are secretly using ChatGPT to help with diagnosis” to plenty of teachers and students using ChatGPT to either make, grade, or otherwise help with school, it’s really time to begin integration.

The word “integration” is a very specific term of art, here. Think about the internet, it was not immediately or obviously incorporated into healthcare or education. Platforms like EPIC and Blackboard became possible with the internet, but it took time for folks to figure out how to build those tech platforms, and get them accepted by their target industries.

It would be my (and plenty other people’s) preferences to just burn down the current systems wholesale and start from scratch, but realistically, that’s not going to happen. First, the only time “burn it down and start over” has happened in history typically occurs with major regime shifts, like socialist and communist revolutions, or the overthrow of a monarchy. Barring some changing of the guard, the current system will have powerful advocates—unions and lobbyists and billionaire stakeholders—that will dig in and prevent rapid change.

Beyond that, there will be regulatory inertia and expert skepticism. There are few people more skeptical of new technology (and methods) than educators and doctors. Over the past few years, I’ve interviewed plenty of both, and they all say the same thing; if you don’t convince the key stakeholders to use it, it’s DOA. For instance, getting many doctors to switch to EHR was like pulling teeth, and some doctors even chose to go out of business rather than join the 21st century. Likewise, teachers have been extremely skeptical of AI on several dimensions—it breaks grading schemes, it might replace them, and so on.

This is where I prefer to look at each industry in terms of first principles. Most people tend to get bogged down by examining the system as it is today, and mistaking the current arrangement as some sort of planned, intentional, optimal solution. If the current education and health system were deliberately designed, then I’d like a word with the architect!

What is the point of healthcare? What is the point of education?

During one interview, I asked a doctor to define the point of healthcare. He sort of stammered through a few ideas, like “provide care” and “cure disease.” He wasn’t wrong, but the fact that someone who’d been in the profession for decades couldn’t clearly articulate the value-prop of his entire industry was concerning.

“The purpose of medicine is to achieve the best possible health outcome for the patient.” ~ My definition of medicine

So here’s the first departure. Doctors have things like the Hippocratic Oath and then a bunch of behaviors like “intake” and “assess” and “design a plan of care” but if the entire healthcare industry hasn’t even clarified or defined its purpose, articulating in crystal clear terms like this, of course no one is going to understand how to integrate AI!

Here’s why; every stakeholder has a different agenda. For-profit hospitals? Make money. For-profit insurance companies? Make money. Individual doctors? Who knows. Some of them want prestige, others want money, and yeah, some of them care about their patients first and foremost.

Now, you might think “Dave, those are just words, who cares?” but let’s reframe. If “achieve the best outcome for the patient” becomes the primary objective, around which we can design, develop, deploy, and integrate AI, then it becomes very, very clear where, when, and how we use AI.

Here’s a screenshot of my chronic illness recovery dashboard that I built with Claude Projects.

I use a combination of KB articles, logs, conversations, and Claude Styles (personalities) to help with tracking my health, learning what I need to about recovery, and documenting everything. This required me to be an expert in AI (to write all the custom instructions and styles, shape the project) as well as pretty savvy with my own health. Claude is smart, and with the right prompts, it can really untangle medical mysteries, but I was bringing a lot of my own intelligence and learning into the equation. This kind of service simply doesn’t exist. Up until now, unless you had millions to throw at personal doctors, it was not possible to get this level of personal care. The best you could do was band together with similar people on the internet, and we all know how reliable anonymous strangers are on the web.

Here’s the thing: AI works really well if you give it clear instructions. That’s why achieve the best possible health outcome for the patient is important.

You know where I got that idea from? The notion of perfectly enumerating the mission of doctors? Lawyers. You see, lawyers are wordsmiths. Legalese might seem byzantine, but it’s really quite a sophisticated rhetorical system, to express exactly what you mean as clearly as possible. Do you know what oath lawyers take? What they are legally required to do in order to be lawyers?

“To zealously advocate for the interests of their client.” ~ The mission of lawyers

You know what the original word for “lawyer” was? Avvocato. No, not “avocado”—avvocato means “advocate” in Italian, while this comes from the Latin advocatus—or “one who is called upon.” Now, think about every time you’ve interacted with a lawyer, with them on your side of the table or otherwise. Everything they do orbits around that mission. It’s why lawyers will sometimes interpret or twist the facts or the law to suit the narrative of their client, because failure to do so constitutes a breach of their responsibilities. You’re literally paying the lawyer to be your zealous defender.

Before we move on to teaching, I need to point out that what I’m going through right now is what ancient Chinese philosophers would call the rectification of names. To the Eastern mind, semantic and rhetorical clarity should precede all action and decision. In the West, we’ve sort of reverse engineered this practice, particularly when we say “Well, it depends on what you mean by [X]?” There are significant geographically and cultural differences that led to these views. The TLDR is that us Westerners have been sea people and merchants for a long time, which means we’re accustomed to negotiating meanings across linguistic and cultural barriers. The Chinese? Not so much.

So, what is the purpose of education?

That really depends on who you ask. Is it to get good grades? To prepare docile workers? To make children ready for life? To give them valuable skills? Or is it to indoctrinate them into their nation’s values? To ensure they replicate their cultural context as adults?

Here’s my distillation of education:

“The purpose of education is to help students reach their maximum potential in life.” ~ My distillation of education

Think about it—what are all those grades and learnings for? Why learn math and history and engage in physical education? Why learn social skills and a work ethic? Why learn how to learn at all? It all orbits around “achieving your maximum potential” and grades serve several purposes; comparison trajectory. We tend to penalize children for bad grades, but maybe that’s just what their best is? It also serves for us to understand who’s doing well—who’s taking to their courses “like a duck to water” versus who are “flying like a brick.”

So what if we reframe AI not as “disrupting the classroom” or “breaking grades” but instead look at it as “how can we use AI to help students achieve their maximum potential?” The conversation is very different there.

Likewise, if we ask “How can AI help patients achieve the best possible health outcome?” Again, we end up with a very different conversation than if we center doctors and hospitals. Medicine is not about doctors, it’s about sick people. Education is not about teachers, it’s about students. Catch my drift? Even law is not about lawyers, it’s about clients.

Sure, no one exists in a vacuum. You can’t have students without some form of teacher, but there’s no law of physics that says it must be a particular kind of human. Likewise, you wouldn’t have doctors without patients, but again, there’s no cosmic mandate that “doctors must be godlike humans in a lab coat.”

My goal with this blog post is to focus the conversation and appeal to first principles thinking as we move into this new phase of integration, consolidation, and start to incorporate AI into civilization.