Sophia Dew interviews Andrei Georgescu, co-founder and CEO of Vivodyne, about a platform that grows patient-derived human tissue and tests drugs and genetic changesAI drug discovery applies AI models to tasks such as target selection, molecule or protein design and experimental prioritization during drug development. in automated labsAn automated AI laboratory organizes research so AI agents execute substantial experimental workflows under human direction.. Andrei Georgescu argues that experiments on human tissue can reveal safety or efficacy issuesExperimental validation tests a prediction or proposed mechanism with measured observations from a suitable real-world or laboratory experiment. earlier than animal studies and before a candidate reaches a clinical trial. The accuracy figures discussed are company-reported results rather than independent clinical validation.
Andrei Georgescu explains why scale matters: robotics can run millions of tissue experiments a yearAutomated experiment execution uses software or agents to configure, run and monitor an experiment according to a defined procedure., including combinations of drug targets and tests that could not ethically be performed in people. He describes work with pharmaceutical companies but does not identify a specific clinical trial, so those examples remain company-reported case studies.
Andrei Georgescu then connects the lab to AI. He describes tissue experiments as an environment for reinforcement learningReinforcement learning trains an AI system to choose actions using feedback about the results of its earlier choices., where rich measurements of diseased and healthy tissue could guide the next intervention to test. He argues that biological AI models need experimental feedback to move beyond recognizing patterns, while combining robotics, software and physiology remains a difficult practical challenge.
Watch on YouTube
