"To play a wrong note is insignificant; to play without passion is inexcusable."
Conceptual
Aligning intelligent, agentic systems to human values is misguided, and we ought instead focus on incorporating the rich computational structure of metaphysical objects, based on insights from moral epistemology.
Applied
Infrastructure for coordinating safe generative deployments on a global scale is still in its infancy. I'm grateful to be part of an organization that's working to change that. We're currently addressing misuse risks arising from dual-use capabilities as a warm-up, with the real challenges being yet to come.
Acknowledgements
I'm grateful for having been supported in my work, either financially or through compute, either recently or less so, by Open Philantropy, Long-Term Future Fund, Google Research, Conjecture, Stability AI, AI Safety Camp, and Andreas Stuhlmüller.