Midjourney’s Medical Pivot and the One Question Every Product Team Should Ask

The news landed on a Thursday, which is the only reason I double-checked before writing about it. Not April 1st. Not a parody. Midjourney — the company that made the AI image generator — has announced Midjourney Medical and a full-body ultrasonic scanner.

Sixty seconds. Full-body scan. Visual quality comparable to MRI.

I want to walk through what they’re building, how they got there, and what I think product managers should actually take from this story — because the obvious take (“AI companies are pivoting to hardware”) misses the more useful one.

What the scanner does

You step onto a platform, get submerged in water, and pass through a ring of half a million sensors the size of a grain of sand. Each sensor emits ultrasonic waves and records the echoes coming back from your body. Midjourney describes it as echolocation — “like being surrounded by half a million tiny dolphins from every angle.”

The output is a 3D map of your body accurate to a fraction of a millimeter, with visual quality comparable to an MRI — in under 60 seconds versus the 60 to 90 minutes standard MRIs require.

They developed this with Butterfly Network, a company that makes ultrasound-on-chip technology, through an exclusive licensing deal signed in November 2025. The hardware effort is led by Ahmad Abbas, who joined Midjourney in late 2023 after working on the Apple Vision Pro.

Eighteen months of preparation. While everyone was focused on their image generation products.

The positioning

The technology is genuinely interesting. The framing is more interesting.

Midjourney describes their vision as: “We’ve dreamed of something as powerful as MRI, and as casual as a trip to the spa.”

That second clause is doing a lot of work.

MRI carries psychological weight: a large, enclosed machine, lying still for an hour, loud, uncomfortable, and you’re anxious about what they’ll find. Most people don’t get MRIs proactively — they get them when something is already wrong. The experience is medical in the most clinical, intimidating sense.

Midjourney is explicitly positioning the Scanner as a spa experience: water, 60 seconds, a San Francisco location that sounds like a wellness center. Same diagnostic capability, completely different emotional context.

This is product positioning, not marketing copy. The decision to use a spa model rather than a clinic model changes who the product is for, how often they’ll use it, and what it costs to acquire them as customers.

What this tells us about AI companies

There’s a common mental model of “AI companies” that treats the category as a destination. You build AI, you are an AI company, you serve AI use cases.

Midjourney is suggesting a different model. Their actual competency isn’t AI images — it’s processing visual data and building consumer-facing interfaces around complex technical systems. That competency generalizes. Medical imaging is visual data. The interface challenge is the same: take something technically dense and make it accessible to someone who isn’t a specialist.

Amazon started with books because Jeff Bezos wanted to build the “everything store” and books were the easiest category to start with — not because Amazon was a book company.

The question worth asking isn’t “what AI company will pivot to healthcare next.” It’s “what specific competency does your team have, and where else does that competency apply?”

The roadmap

Midjourney’s timeline: algorithm refinement through 2026, first Spa location in San Francisco in 2027, FDA approval process underway, third-generation hardware with custom silicon in 2028, expansion to more cities. Goal: 50,000 scanners by 2031.

Their stated ambition: “with enough early imaging, the world could avoid 30 percent of all deaths and 50 percent of all healthcare costs.”

That’s an enormous claim. It’s also the same kind of claim that sounded unrealistic when private companies first said they’d reach orbit.

Two questions for product managers

After reading about this, I found myself thinking about two questions that don’t come up enough in product work.

First: what does your team actually do? Not the product you ship — the underlying competency you’ve developed. If you had to apply that competency to a completely different problem, what would that problem look like?

Second: where in your market is something being avoided because it feels too clinical, too technical, too intimidating — when the underlying capability is genuinely useful? Midjourney found that gap in healthcare. Most industries have one.

The answers to both questions are worth more time than most product teams give them.

More on unexpected AI company pivots in my Telegram channel PM vision.