Artificial Intelligence Accidentally Learned Ecology through Video Games

Lou Barbe; Cendrine Mony; Benjamin W Abbott

doi:10.1016/j.tree.2020.04.006

Artificial Intelligence Accidentally Learned Ecology through Video Games

Trends Ecol Evol. 2020 Jul;35(7):557-560. doi: 10.1016/j.tree.2020.04.006. Epub 2020 May 7.

Authors

Lou Barbe¹, Cendrine Mony², Benjamin W Abbott³

Affiliations

¹ ECOBIO, OSUR, CNRS, Université de Rennes 1, 35000 Rennes, France. Electronic address: lou.barbe@univ-rennes1.fr.
² ECOBIO, OSUR, CNRS, Université de Rennes 1, 35000 Rennes, France.
³ Department of Plant and Wildlife Sciences, Brigham Young University, Provo, UT, USA.

PMID: 32389438
DOI: 10.1016/j.tree.2020.04.006

Abstract

An advanced artificial intelligence (AI) system defeated the best human players in StarCraft II, a popular real-time strategy game. In a virtual ecosystem, players compete for habitats and resources, unintentionally reproducing many ecological phenomena. We propose to repurpose this AI to test ecological hypotheses that have been intractable using traditional approaches.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Artificial Intelligence
Ecosystem
Humans
Learning
Video Games*