Amica Conference:
Thermodynamic Approaches to Artificial Intelligence

The Amica Conference: Thermodynamic Approaches to Artificial Intelligence seeks to foster interdisciplinary dialogue on the foundational ties between AI and thermodynamics. On the physics side, stochastic thermodynamics has provided a powerful framework for clarifying the relationship between information and energy, while on the AI side, modern generative models—including the diffusion model—have drawn inspiration from thermodynamics and statistical physics. By hosting a small, focused gathering of leading researchers, the conference aims to illuminate emerging methodologies and chart new directions for research at the intersection of statistical physics and machine learning.