Summary
David Dalrymple is a programme director and interdisciplinary researcher with 15 years building systems at the intersection of AI safety, applied category theory, systems neuroscience, and programming languages. He currently leads R&D funding at ARIA to create explainable mathematical models of real-world phenomena using large language models, after co-inventing Filecoin and inventing Hypercerts at Protocol Labs. His background ranges from hands-on production engineering at Twitter and Magic to academic research at FHI, MIT Media Lab, and Harvard, culminating in the Open Agency Architecture for technical AI safety. He has founded and coordinated ambitious lab projects in computational neuroscience (Nemaload) and has deep experience shipping high-throughput, resilient infrastructure and novel proof-of-concept research. Known for bridging rigorous math and practical engineering, he combines a rare mix of formal modeling instincts with production-grade systems design. Based in the UK, he brings both entrepreneurial grit and academic depth to hard problems in trustworthy AI and mechanistic modelling.
16 years of coding experience
12 years of employment as a software developer
Mesh, Mesh at Recurse Center
S.M. Media Technology, S.M. Media Technology at Massachusetts Institute of Technology
UMBC
Doctor of Philosophy (Ph.D.) Biophysics, Doctor of Philosophy (Ph.D.) Biophysics at Harvard University
Artificial Intelligence Networks and Computing Systems Neurotechnology, Artificial Intelligence Networks and Computing Systems Neurotechnology at Singularity University