On AI:AM

David Duvenaud

Professor of Computer Science, University of Toronto; founding faculty, Vector Institute

David Duvenaud is a professor of computer science at the University of Toronto and a founding faculty member of the Vector Institute, known for foundational machine learning work including Neural ODEs (a NeurIPS 2018 best paper) and the autograd automatic-differentiation library. After an extended stint working on AI alignment and dangerous-capability evaluations at Anthropic, he turned his research toward what happens to civilization after AGI. He is a co-author of 'Gradual Disempowerment,' which argues that even well-aligned AI could erode human control of the economy, culture, and the state — not through any takeover, but as institutions stop depending on human labor, votes, and attention — and he convenes the Post-AGI workshop series searching for a stable, good post-AGI equilibrium.