This lecture by the instructor covers advances in interatomic potentials for materials, focusing on multi-scale modeling, fitting functions in high-dimensional spaces, and applications of machine-learned interatomic potentials to solve long-standing materials problems. The lecture delves into the accuracy and qualitative aspects of multiple scales of materials modeling, the transferability and accuracy of interatomic potentials, and the challenges in creating accurate, reactive potentials for covalently bound materials. It also explores the representation problem of atomic environments, atomic neighbor density functions, and the Smooth Overlap of Atomic Positions (SOAP) method. The lecture concludes with discussions on the performance comparison of various interatomic potentials and the future perspectives for achieving meV accurate force fields.