There are things you don't see coming. And Midjourney just delivered one of them.
If you follow the world of artificial intelligence, you have probably come across stunning images — floating castles, hyper-realistic portraits, scenes that look straight out of dreams — generated in seconds by the platform. Midjourney became the symbol of an era: that of machines learning to create.
But on June 17, 2026, CEO David Holz took the stage in San Francisco and announced something that left even the most plugged-in people in the field scratching their heads.
Midjourney's new bet has nothing to do with art. It has to do with you — or rather, with what is inside you.
The company announced Midjourney Medical, a new health-focused division whose first product is the Midjourney Scanner: a full-body ultrasound machine capable of generating detailed images of your body in about 60 seconds.
No radiation. No magnetic fields. None of that claustrophobia typical of an MRI.
The process is almost meditative: you lie on a platform that lowers slowly — at just 5 centimeters per second — into a tub of warm water. "You go into the water, you come out of the water, and that's it," the company describes. In less than a minute, you already have a visual map of your muscles, fat, bones, and internal organs.
Midjourney didn't build everything from scratch — and that, in fact, is part of what makes the project credible.
In November 2025, the company signed an exclusive licensing deal with Butterfly Network, a reference in ultrasound-on-a-chip technology. Each scanner uses 40 modules of that technology, supported by a processing capacity of two petaflops — enough to turn the chaos of sound waves into comprehensible images.
An honest caveat is in order: the current resolution is still lower than that of a conventional MRI. Midjourney itself acknowledges that the initial product is not a clinical diagnostic tool. For now, the focus is on body-composition maps — the kind of data that personal trainers and sports physicians love, but that is rarely accessible to the general public.
The path to deeper diagnostics runs through FDA approval, and the company has already announced it will work on that progressively.
This is where things get really interesting — and a little cinematic.
Midjourney doesn't want to sell its scanners to hospitals. It wants to open spas.
Yes, you read that right.
The plan is to create luxury wellness centers — with saunas, cold plunge pools, and relaxation rooms — where the scanners are available as part of the experience. The idea, in the company's words, is for "the scans to be a side effect" of a visit to the spa.
The first space is expected to open in Union Square, San Francisco, in late 2027. The long-term goal is ambitious: 50,000 scanners around the world by 2031.
Holz doesn't think small. In his presentation, he said that "within 10 years, these devices won't just be imaging diagnostic tools — they will probably also have therapeutic applications."
It's a bold claim. But it comes from someone who already showed the world that images can be generated by algorithms — something that also seemed impossible until it was demonstrated live.
What Midjourney is doing here strongly resembles the move companies like Fujifilm made decades ago: realizing that the true asset is not the product itself, but the know-how accumulated in image processing — and applying it to new contexts.
For 2028, the company plans to expand to other cities and launch a third generation of hardware with custom silicon, when, according to them, things will get "serious" in terms of competition with traditional medical equipment.
It's fair to keep both feet on the ground. There are still no independent clinical validations. No external radiologist has publicly assessed the quality of the images. Regulatory approval is a long and uncertain road.
But it's also fair to recognize: the idea of democratizing access to body imaging — making health monitoring as natural as going to the gym — is genuinely powerful.
If Midjourney manages to deliver what it promises, the question stops being "is this possible?" and becomes "why didn't we do this before?"
And maybe that's the best definition of what artificial intelligence has been doing for the world: not just answering old questions, but making us see new ones.
Sources: Bloomberg, The Verge, Engadget, Butterfly Network
