Medvi, glorified as a two-employee, $1B+ startup, becomes a cautionary tale about AI-enabled hype

Background

It has been reported that the New York Times ran a glowing profile of Medvi that portrayed the company as a two-person startup claiming more than $1 billion in revenue. The item spread quickly through Techmeme and social feeds, presenting a tidy, almost cinematic image: tiny team, outsized success. The story landed like a neat little Silicon Valley fairy tale — until questions started piling up.

The unraveling

Gary Marcus, writing on Substack, alleges that much of Medvi’s public narrative was engineered with the help of generative AI: website copy, testimonials, partner claims — even aspects of the company’s provenance — it has been reported that these elements may have been exaggerated or fabricated. Investors and reporters, it has been reported, picked up the story and amplified it without deep verification. Allegedly, what looked like a brilliant, efficient startup was at least partly a marketing construct, not the fully documented business it was billed as.

Why it matters

So what’s the lesson here? AI is a force multiplier for creation — and for deception. When tools can spin convincing narratives in minutes, sloppy fact‑checking and PR theater get a dangerous assist. How did this slip past journalists and investors? Good question. If the allegations hold up, the emotional sting is straightforward: trust betrayed. The fix is not just better tools; it’s harder-nosed verification, smarter reporting, and a willingness to ask uncomfortable questions before the applause starts. No one likes a con, even one dressed in state‑of‑the‑art code.

Sources: garymarcus.substack.com