Medical superintelligence is here, but Is humanity ready?

Welcome to AI Collision 💥,

In today’s collision between AI and our world:

Microsoft’s AI diagnostic god
Playing it down, because of its power?
How would you feel with AI telling you the outcome?

If that’s enough to get the healthcare revolution underway, read on…

AI Collision 💥

A few weeks ago, I found myself with a sharp pain in abdomen while watching my son’s karate class.

The pain ratcheting up by the minute. It wasn’t like a pain I had before, so, I did what any logical, mildly panicked, tech-savvy human in 2025 would do… I asked ChatGPT.

Not for entertainment. Not for trivia. But for a medical opinion.

I fed it my symptoms and asked to give me a checklist. And what I got back was as close to a reasoned, accurate differential diagnosis as you’d expect from a competent doctor… with a little dose of urgency too.

It said, “Get to the hospital emergency ward, NOW!” I did. And it was right. Appendicitis and a stint in hospital that weekend was the outcome.

So, when Microsoft posts a calmly worded announcement titled “The Path to Medical Superintelligence”, I don’t read it like it’s some future fantasy. I read it as someone who knows firsthand we’ve already crossed the threshold.

Microsoft dropped a proverbial AI bomb on the world this week, and I don’t think it’s quite got the attention it deserves.

Microsoft is building a diagnostic god.

In typical corporate understatement, Microsoft frames their new system as an early research exploration. That’s despite it outperforming trained clinicians on New England Journal of Medicine (NEJM) case studies – a gold standard for real-world medical complexity.

“Microsoft’s MedPrompt ensemble system answered 80% of them correctly. Doctors? Just 2 out of 10.”Benchmarked against real-world case records published each week in the New England Journal of Medicine, we show that the Microsoft AI Diagnostic Orchestrator (MAI-DxO) correctly diagnoses up to 85% of NEJM case proceedings, a rate more than four times higher than a group of experienced physicians. MAI-DxO also gets to the correct diagnosis more cost-effectively than physicians.

A colleague of mine put it bluntly after he’d read the link I posted on Wednesday about all this,

“Microsoft is downplaying the impact massively… if AI can do as well as Microsoft claim there is no way people will go to in-person clinic only to get “trust and empathy”

I agree wholeheartedly.

If an AI can now read symptoms, suggest tests, interpret x-rays, and out-diagnose doctors, why on earth would people bother with the infuriating process of trying to get a clinic appointment or even seen by a medical professional inside of 12 months, when the best they’ll get anyway is not as good as AI can deliver?

This is what disruption looks like, but it’s also a Catch-22.

There’s a reason Microsoft hasn’t pushed on this too hard. And my view is they’re acutely aware of the significance of this and know they can’t go too hard with this too soon.

Tens of thousands (maybe millions) of medical professionals, especially in diagnostic roles, will soon face obsolescence. And yes, it could be very much in the “overnight” sense… if the full power of what Microsoft have proven is unleashed.

But, that’s not going to be allowed now, is it?

It’s clear this could result in clearer and better patient outcomes. And there’s an added benefit of cost savings in healthcare, which for governments should be a godsend. We all know healthcare systems like the NHS are stretched as it is, surely when something like this is available, you’d move heaven and earth to implement it?

Imagine the benefit to the public if the cost and access for diagnostics improves ten-fold (or better).

But then again… that displaces a lot of professionals in a reasonably specific occupation. And fewer jobs mean fewer tax receipts. Fewer workers mean fewer voters supporting bloated healthcare bureaucracies. If innovation guts the system too fast, the very thing that was supposed to save the system becomes the thing that breaks the government’s control levers.

And in response? More regulation. More taxation. More control.

This is the cycle. We invent solutions to make life better and then panic when the world around them begins to collapse under their potential. And we end up with controlled, siloed and restricted technologies for the few and not the many.

We saw it with automation in manufacturing. We’re seeing it now in AI software development. And we’ll see it again (perhaps more brutally) with healthcare.

Then come the robots.

Let’s not forget the other AI elephant in the room, humanoid robotics.

Where do you think the first large-scale deployment of general-purpose humanoid robots will go? Probably not logistics. Not restaurants. It’ll be hospitals.

Why? Because they’re environments full of repetitive, critical tasks, lifting, fetching, dispensing, even providing company and comfort for patients. And hospitals are desperate for more labour.

As the diagnostics shift to AI, the rest of healthcare won’t be far behind. You’ll get scanned by a robot, assessed by an LLM, treated by another machine, and perhaps discharged faster than ever. You may not ever need to see a human doctor or nurse… or the ones you do will have the time, and stress-free existence to actually give you the full attention you and every other patient deserves. Heck maybe AI and robots mean you get your own private human nurse the whole time you’re in hospital.

It’s not sci-fi. It’s simply progress.

But again, it’s a Catch-22.

All this progress – as brilliant and inevitable as it is, faces one massive bottleneck, humans.

Humans are the greatest threat to human progress. Not because we’re evil. But because we panic, we worry, we regulate, we politicise.

Humans don’t like massive change unless it’s comfortable, slow, and full of familiar faces.

But the technology doesn’t care.

So now we’re at a fork in the road. We can either lean in, reimagine healthcare, retrain at scale, build new support systems, and embrace medical superintelligence… or we can try to bottle it up, tax it, control it, and pretend we don’t need it as desperately as we do.

One path leads to cheaper, faster, smarter healthcare for billions.

The other leads to stagnation and more of the same old nonsense we’re used to with governments and healthcare.

AI isn’t the problem. Innovation isn’t the enemy. Progress doesn’t break the world, we do, when we try to control it.

So yes, Microsoft might be downplaying it but only because they know how powerful it really is.

The future of healthcare is already here, and it has real-world potential to save lives.

Deep Blue Scientist Reveals Hidden AI 2.0 Breakthroughs

James Altucher – a 40-year AI veteran who helped build IBM’s Deep Blue – says most people are looking at the wrong AI tech.

In a new video, he exposes why flashy tools like ChatGPT are just the tip of the iceberg –and reveals a handful of under-the-radar AI 2.0 companies quietly leading a $15.7 trillion revolution.

Click here to see his full breakdown.

Capital at risk.

Boomers & Busters 💰

AI and AI-related stocks moving and shaking up the markets this week. (All performance data below over the rolling week).

man in black suit jacket and black pants figurine

Boom 📈

Hewlett Packard Enterprise (NYSE:HPE) up 16%
Everspin Technologies (NASDAQ:MRAM) up 8%
Apple (NASDAQ:AAPL) up 6%

Bust 📉

Micron (NASDAQ:MU) down 3%
Tesla (NASDAQ:TSLA) down 3%
AMD (NASDAQ:AMD) down 4%

From the hive mind 🧠

I wrote to you this week about the use of AI in music and how it was quite possibly an evolution in music, not the death of it. Here’s another take on it, but from someone who’s actively using it and says it’s “amazing”.
There will be pushback on the use of AI in healthcare and medicine, that much is to be expected. And I think it’s going to be a massive battleground for future election campaigns.
If you can’t buy them… then just buy all their employees.