“Julian Francis Miller” Awardee Keynote

Endless forms most beautiful: quality, diversity, and the MAP-Elites algorithm

The dominant theoretical framework for artificial evolution has long been, and to a large extent still is, single-objective black-box optimization. This view sidelines one of the most striking and enigmatic feats of natural evolution: its ability to autonomously generate an ever-expanding variety of refined yet disparate lifeforms. Put differently, it focuses on micro-evolution, and ignores macro-evolution. In this talk, I will trace the major milestones in our quest to design algorithms that capture some of this generative power while remaining simple and practical. This family of algorithms is now known as « quality-diversity algorithms ». I will illustrate the talk with recent results from diverse fields, from chemistry to robotics, and discuss the main open challenges and opportunities .

Jean-Baptiste Mouret

Jean-Baptiste Mouret is a senior researcher (“directeur de recherche”) at Inria (Nancy, France) and the co-founder of Bleu Robotics. His research intertwines evolutionary computation, data-efficient machine learning and robotics to design creative machines that adapt and learn from their experience. He is one of the architects of the quality-diversity algorithm family, including MAP-Elites. His work was featured on the cover of Nature (2015) and funded by an ERC Starting Grant (ResiBots). He received several awards, including the “Prix La Recherche” (2016) and the “Distinguished Young Investigator in Artificial Life” (2017).

More information about the “Julian Francis Miller” Award at https://species-society.org/julian-francis-miller-award/