AI Risks and what mathematicians can do about them

Abstract: This talk aims to give an overview of the state of AI progress and scaling laws, the risks (both immediate and long-term) that human civilization faces. These risks range from the misuse of AI capabilities in cyber, bio, and persuasion, to the inexorable trend towards gradual disempowerment, to existential risks of full AI takeover and extinction. I will lightly survey progress on the three major directions that we can contribute to as mathematicians and individuals: theoretical AI alignment, empirical AI safety, and AI pause advocacy.

machine learningmathematical physicscombinatoricscategory theory

Audience: learners

( paper )

Tropical mathematics and machine learning

Series comments: Tropical mathematics, machine learning, category theory and anything tech+math are welcome.