Accelerate23 - AI will be inseparable from computational materials science

A talk on how my views on how best to approach computational materials science over the past decade.

Key URLs:

Abstract: Artificial Intelligence methods provide capabilities widely distinct from and complementary to existing computational methods. This presentation will cover how practices in computational materials engineering that were static for decades have evolved to leverage the wide availability of flexible, intelligent algorithms. Empirical models that were permanently fixed now evolve on-demand, and design processes that used to have humans or automation at the center are now built around creative software. Numerous other changes make a future where “computation” and “data science” remain separate unlikely.