It starts by stepping into a shallow pool of golden light. You then begin to descend into the water. Your body passes through a ring of underwater sensors, each acting like a dolphin, using its echolocation. The sensors send ultrasonic sound waves through your body from every angle. With enough waves, and enough angles, we form an image of what’s happening inside your body.
It sounds like it’s taken directly from a 1980s sci-fi script, but it’s in fact the description of the latest innovation from Midjourney, the company that first launched their image generating service on Discord in 2022.
The idea is a sort of new kind of MRI that is both faster and a lot cheaper than the large and expensive machines used today. This is achieved through the use of ultrasound fired off against your body by 500,000 tiny receptors, while under water. Midjourney claims that this will create terrabytes of data for every second.
As the ultrasound waves bounces off your body, Midjourney’s special AI converts them into high-resolution images and turns them into a 3D map of your body. Both the inside and outside. The first scanner is planned to be ready by the end of 2027, and by 2031 an estimated 50,000 scanners will be available worldwide.
These scanners will be able to perform a billion scans per month, and will be cheap enough for most people to afford.
Unintended Effects
How will this affect our healthcare services?
To make an analogy, 100 years ago a country like Sweden, located in the very north of Europe, could have a complete open border policy. Because 100 years ago most people wanting to migrate did not have the physical capability to travel to a small country in the north.
The reality today is that movements are very different, and a completely open border today would create chaos.
Today, not everyone who wants a full body scan (in the form of an MRI) can have one, and the healthcare services are able to handle those who do get a scan.
When you remove the friction of cost and accessibility from a diagnostic tool, you don’t necessarily eliminate scarcity; you just shift the bottleneck downstream. If a full-body scan becomes as cheap and easy as getting a haircut, the immediate impact on healthcare services wouldn’t just be a strain, it would likely be a systemic shockwave.
- The Anxiety Loop: If millions of asymptomatic people suddenly get full-body scans, hundreds of thousands will discover they have a “minor shadow” or a “micro-nodule.”
- The Downstream Flood: A cheap AI scan might point out a spot, but it can’t perform a biopsy. Patients will panic and flood primary care physicians, demanding specialist referrals, blood tests, and confirmatory MRIs.
Proponents argue that early detection saves money. While catching a disease early is clinically superior, mass-screening an entire healthy population will likely have the opposite financial effect.
Explosion of Healthcare Spending: Even if the initial scan costs $50, the subsequent diagnostic odysseys (biopsies, follow-up scans, specialist consultations) are incredibly expensive. In publicly funded systems like Sweden’s, or insurance-based systems like the US, this could trigger severe budget crises.
Countermeasures
Healthcare systems will have to build incredibly strict regulatory “gatekeepers” to decide what AI findings are actually allowed to trigger human medical intervention.
Even if it were to explicitly say “You have a 2mm shadow on your liver, but based on 10 million data points, there is a 99.9% chance it is benign. Do not go to a doctor.”, it won’t be effective.
People are easily scared, often feel entitled, and believe that their specific needs should have precedence before everyone else. If primary care doctors refuse to see these patients because the AI cleared them, people will simply find the path of least resistance to a human face.
People will flood the Emergency Departments, hand the triage nurse a PDF report from Midjourney, and claim they are having chest pains or shortness of breath just to force a human medical workup. So could we prevent this simply by making new laws to regulate the way Midjourney and other AI companies handle these scan images? Prevent them from showing “benign background noise” from the scans?
I believe it is highly unlikely we will see a law like this, for a very simple reason. Any proposal for a law like this will be public, and the proposal alone would create an outcry. Politicians would be accused of not caring about the people at best case, or accused of wanting to kill the population in the worst case. The opposition would promise that, if they get elected, they will abolish the law.
It would be political suicide for anyone to suggest such a law.
The political reality is that the data will be given to the public. But instead of a smooth, AI-managed utopia, we are likely heading toward a bifurcated system: a Wild West commercial market fueling high-resolution anxiety for anyone who can afford a spa ticket, constantly crashing against a rigid, defensive traditional healthcare system that refuses to validate the findings.
Independent research like this is self-funded. If this guide saved you hours of troubleshooting, consider fueling the lab. You can now make completely secure one-time or monthly contributions through Google.
The Middleground
Can we count on AI companies being responsible and hiding certain things from the scans, without the involvement of politicians and healthcare providers? Highly unlikely. This is the type of information that would leak pretty quickly via a disgrunted employee. Alternatively, a rival company will offer the full results from their scans.
Let’s say the market leader, “Company A,” acts responsibly. They decide their AI will quietly suppress benign data to avoid triggering a public health panic.
It would take about five minutes for a venture-backed competitor (“Company B”) to realize they can capture 100% of the anxious market by launching an aggressive marketing campaign: “Unlike our competitors, we don’t hide your data. We give you the UNFILTERED TRUTH about your body.”
If it were up to me, I would make a policy that looks something like this:
We don’t acknowledge wellness spa scans as a base for medical attention. The company that makes the scans will have to evaluate them, hiring their own doctors to do so if necessary.
Whether the company wants to pay for its doctors out of its own pocket and offer that service for free to its customers, or charge them directly for the doctor’s review, will be up to them.
At least for the time being, these scans are referred to as part of a “spa” treatment. While Midjourney states that they continuously will submit results to the FDA, and I assume the eqvivalent regulatory agencies in Europe, it should be possible to add a clause in an eventual approval – simply stating that they will have to manage the needs that directly occour from their scans.
By telling the company, “Your scan, your doctors, your problem,” the state completely protects its primary care infrastructure. If a consumer gets a scan and wants a professional opinion on a weird shadow, they have to pay the spa’s in-house private physicians to look at it. The cost of paying for private doctor consultations to review every minor aesthetic anomaly will naturally deter the “worried well”.
If the licensed wellness-spa doctor looks at a scan, meets the patient, and decides that further care is needed, referral to public healthcare is completely legit. When a patient finally does transition from the private spa to the public hospital, they are no longer an anxious consumer wielding a confusing 3D printout from an app. They are a vetted patient with a formal, peer-to-peer clinical referral.
Scans only for the rich then?
No, I believe any AI company could still offer cheap scans for anyone who wants them. But until their own doctor has reviewed the scans, they are nothing more than a funny gimmick to hang on your wall. By keeping the scan cheap but paywalling the medical interpretation, the market naturally resolves the chaos through a classic “Freemium” business model.
Wouldn’t people just look at their own scans and see a strange shadow or structural anomaly?
Most people don’t know where their liver is located, what the purpose of a kidney is, or what the difference between a gall bladder and pancreas is. Handing a raw, high-resolution 3D anatomical file to the average person is like handing the engineering blueprints of a fighter jet to someone who just figured out how to check their tire pressure.
The Current Reality
The current reality is that we probably don’t have to figure out how to prevent hordes of healthy people showing up at our hospitals, demanding confirmation through biopsy and blood tests. At least going by the images available on Midjourney’s website.
I assume that the actual text won’t be in the final scans, as human anatomy isn’t a standardized factory blueprint. Our blood vessels branch differently, our muscle densities vary, and our organs shift based on everything from posture to what we ate for breakfast.
Because of how unique we all are, using a static overlay with text is not an option.
If you enjoyed this post and want more in the future, subscribe to my newsletter to ensure you don’t miss an update.