AI isn’t close to curing cancer. This startup says it knows what it will take.
It's the data, stupid.
A biotech startup named Vivodyne claims that the AI drug-discovery industry is grappling with a data issue, and it has developed a solution. The company's modular robotic labs, known as HIVE, can cultivate 20 types of human tissue and autonomously administer dosages and monitor them, generating unique biological data that current AI models lack.
Vivodyne's CEO and co-founder, Andrei Georgescu, emphasizes that without human testing, AI models are prone to predicting cancer cures in mice rather than humans. This sentiment was echoed by Anthropic CEO Dario Amodei, who recently commented on the overused claim that AI could cure cancer. While some AI-designed drugs have made it to human trials, the industry still faces significant challenges, as 90% of drugs that succeed in animal testing fail the transition to human trials.
Vivodyne differentiates itself by claiming that its tissue culture systems closely mimic real human organs, achieving high predictive accuracy for toxicity and other biochemical properties. The startup recently opened what it calls the world's largest "human data center" near San Francisco, boasting twice the throughput of all current animal trials in the US.
Georgescu sees this as a way to accelerate drug candidate development by providing better predictions of which compounds are likely to succeed before costly clinical trials. He views his autonomous biology labs as crucial for generating the kind of causal data needed to train AI models on human biology.
Written by urgent.news from TechCrunch's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
