Midjourney Goes Medical, Skeptics Are Loud
Midjourney announced a medical imaging product targeting a billion full-body scans per month. The HN thread was immediately hostile. The core complaint: Midjourney's entire brand is built on hallucination, on making things up plausibly. That's charming for fantasy art and genuinely dangerous for radiology. The scale claim alone raised eyebrows. A billion scans monthly would dwarf the entire global medical imaging market several times over.
The pattern here is AI companies sprinting into regulated, high-stakes domains before they've demonstrated reliability in low-stakes ones. Midjourney hasn't shipped a product that requires factual accuracy yet, and now it's pitching diagnostic imaging. Commenters weren't just skeptical of the technology, they were skeptical of the stated scale, treating the billion-scan number as either aspirational fiction or a misunderstanding of the market.
The counterpoint some raised: even imperfect AI screening tools could catch things that currently get missed due to cost or access barriers. But that argument didn't land well in the thread, where the dominant concern was liability, hallucination rates, and the specific mismatch between Midjourney's track record and the demands of medical diagnosis.
So what?
If you're building in health AI, the Midjourney announcement will make your fundraising conversations harder, not easier. Investors and hospital buyers are going to be more cautious about brand-new entrants making big medical claims. The flip side: if you already have accuracy benchmarks and clinical validation, the contrast with Midjourney's vague announcement is now a selling point you should use explicitly.