A startup called Nome decided the pharmaceutical industry wasn't going to solve rare diseases so they'd do it themselves using AI and apparently some humans who know chemistry. They take genetic diagnoses from families whose kids have conditions so obscure there's no drug coming ever and try to develop custom treatments. Custom. Like a suit but for keeping your child alive.
The business model runs on desperation. Families get a diagnosis with a name they can't pronounce and zero treatment options because their kid is one of forty people on Earth with the thing. Big Pharma looks at the addressable market and laughs. Nome walks in and says give us your genetic data and we'll see what we can do. They use AI to speed up drug development which in this context means maybe five years instead of never.
This sounds noble until you remember it only exists because the existing system is designed to ignore anyone who can't generate a billion dollars in revenue. Nome isn't disrupting anything. They're running a boutique operation in the gaps between profitable diseases. They're the food truck parked outside the hospital because the cafeteria won't serve anything that doesn't move volume.
The AI part is doing what AI does everywhere else: pattern matching and optimization at a scale humans can't match. Except here the output isn't a chatbot or an ad campaign. It's a molecule that might stop a six-year-old from dying. No pressure on the training data.
Families with rare disease kids now get to add "negotiate with a startup about experimental drug development" to their list of things to do between specialist appointments and insurance appeals. They get hope dispensed in partnership with a company that didn't exist three years ago. The future of medicine is a Series A pitch deck and a prayer.
At least Nome is trying, which is more than Pfizer will ever do for a market of twelve people.
Leave a Comment