As AIs become increasingly capable, they are changing how we learn about life on Earth and empowering us to do biological engineering in new ways. I am working to ensure that this transformative technology is used for improving human well-being and not for harm. I do this as a research scientist at an organization called SecureBio, where I lead the AI x Bio Group.
Topics of our ongoing research
Super-expert biology
How could we robustly measure AI capabilities in biology once they far exceed those of human experts? A key feature of our current approach is that expert humans can reliably verify an experimental outcome even if they could not predict it.
Biological ideation and problem solving
What is involved—theoretically and practically—when scientists generate novel ideas in the life sciences? How can we study that process in a formal way, and how will current and increasingly powerful AIs change the arrival of new ideas?
Tacit, embodied, and social knowledge of scientists
Which aspects of biological expertise—if any—are not well-represented in the training material of frontier LLMs? What does such expertise consist of? How can we reliably measure it?
Benefits and risks at the biotechnological frontier
How can we use AI to empower biologists to discover and engineer beneficial things while minimizing their capacity to cause serious harm?
Please note: This is a personal website that does not include AI and biosecurity-related topics. Instead, you’ll find non-technical snapshots of some of my previous discovery-based research in cell biology, genetics, development, evolution, ecology, quantitative biology, and computational modeling. I have also included a few explorations in custom lab tools, data visualization, and scientific communication.
