AI models are hallucinating cures — can robotic labs fix the data gap?
vivodyne launched 'hives' to grow human tissue for drug testing, achieving 96% accuracy over animals. this solves the data bottleneck preventing ai from actually curing cancer.
vivodyne spun out of upenn to address the data scarcity limiting ai drug discovery, opening a facility near sf to grow human tissue for autonomous testing. the 'hives' system achieves 94-100% concordance with actual human organs, surpassing the 10% success rate of animal testing in clinical trials. backed by $80m from khosla ventures, the startup works with major pharma to de-risk drug candidates before costly clinical trials. by generating high-quality causal data, vivodyne aims to unlock the scaling laws necessary for ai models to master human biology.