← all projects
ARIA Funding Alignment: AI Scientist
AI for Clinical Research
In a future world where pre-clinical drug development can be automated or done in-silico, the barrier to bringing life-saving treatments to patients will be testing them in humans. Clinical trials currently account for ~70% of the R&D time and cost when developing a new drug and require a series of difficult decisions under uncertainty, from choosing which hypotheses to pursue to deciding what study should be run next and why. Encode fellow Will is building multi-agent AI scientist systems to generate, critique, and compare research strategies. Using clinical research as a demanding testbed, the project also asks a broader question: what capabilities and infrastructure are needed for AI scientists to operate effectively in real-world scientific workflows? By developing systems that capture tacit knowledge, coordinate scientific reasoning and evaluate capabilities, this work aims both to support clinical research strategies and lay the foundations for more useful AI scientists.
Philip Torr's Lab
University of Oxford
,
Oxford
Torr Lab
↗
Will holds a PhD in AI for Healthcare from Imperial College London. He developed Steward.ai, an app to help healthcare practitioners prescribe antibiotics appropriately. Previously, he interned at GSK.ai, investigating the reliability of large language models in biomedical applications.
Philip Torr
Professor of Engineering Science; Five AI/Royal Academy of Engineering Research Chair in Computer Vision and Machine Learning
Philip Torr leads Torr Vision Group, which applies deep learning techniques to a wide ranging set of topics. He has been involved in numerous spin-outs as founder or advisor including: FiveAI, Onfido, Oxsight, Eigent, DreamTech, Visionary Machines, CamelAI, as well as working closely with big tech companies like Google, Meta, Apple, Microsoft, and Sony.
