$1.8 billion. That was the headline valuation attached to Medvi, described as the first AI company to hit that mark. Both the number and the phrase "AI company" appear in quotation marks in the coverage - which tells you most of what you need to know about where the story goes.
Gary Marcus, a cognitive scientist who has spent years tracking the distance between AI marketing and AI substance, published what he calls the backstory behind Medvi's rise. The investigation suggests the valuation, and possibly the AI credentials behind it, didn't survive close examination.
The specific details of Medvi's situation are still emerging from Marcus's account. But the pattern the story fits is familiar: a company positions itself in the AI space, raises capital against projected AI-driven value, and earns a valuation that requires substantial optimism about capabilities that may not exist yet - or at all.
Healthcare is particularly prone to this dynamic. AI claims in medicine are difficult for non-specialist investors to evaluate, and the potential upside if the claims are true is enormous. That combination - hard to disprove, plausible upside - makes it easier for companies to raise significant capital against narratives that are more aspirational than operational.
Marcus's broader argument, consistent across his work, isn't that AI doesn't produce value. It's that the claims made by companies and investors routinely run ahead of what the technology can actually do, and that gap gets exploited. Medvi, if the reporting holds, is an example of that gap functioning as a business model.
For procurement teams or buyers evaluating AI tools in healthcare or any other specialized sector: ask to see the AI working on real tasks, not a curated demo. Ask specifically what the AI component does, and whether that functionality would still be called AI if the label didn't affect the price. The word has been valuable enough as a signal that it gets applied to products where it does very little actual work.