Open-endedness: A new grand challenge for AI

Kenneth Stanley (Uber AI Labs | University of Central Florida)

10:10am–10:30am Thursday, September 12, 2019

Location: Hall 2

Secondary topics: Machine Learning, Reinforcement Learning

Average rating:

(4.08, 12 ratings)

Download slides (PDF)

Watch the keynote

We think a lot in machine learning about encouraging computers to solve problems, but there’s another kind of learning, called open-endedness, that’s just beginning to attract attention in the field. Open-ended algorithms keep on inventing new and ever-more complex tasks and solving them continually—even endlessly.

Kenneth Stanley explores how, if you could genuinely program such open-endedness, the longer your algorithms ran, the more interesting and powerful the results they would produce. Endless designs, expanding repertoires of skills, exploding intelligence—these are the rewards of open-endedness. Interestingly, open-ended phenomena are all around—the history of human invention, the unfolding odyssey of art and music, the career of a creative trailblazer, and most dramatic of all, the awesome divergence of natural evolution into all the diversity of life on earth—the open-ended processes can offer an entirely different level of automated creation. Beyond deep learning and beyond the benchmarks of today, open-endedness offers a new path and a new quest for the future.

Kenneth Stanley

Uber AI Labs | University of Central Florida

Kenneth O. Stanley is the Charles Millican professor of computer science at the University of Central Florida where he’s director of the Evolutionary Complexity Research Group, and he’s a senior research science manager and head of Core AI research at Uber Labs. Previously, he was a cofounder of Geometric Intelligence, which was acquired by Uber to create Uber AI Labs. He’s an inventor of the neuroevolution of augmenting topologies (NEAT), HyperNEAT, and novelty search neuroevolution algorithms for evolving complex artificial neural networks. His main research contributions are in neuroevolution (i.e., evolving neural networks), generative and developmental systems, coevolution, machine learning for video games, interactive evolution, and open-ended evolution. He’s won best paper awards for his work on NEAT, NERO, NEAT Drummer, FSMC, HyperNEAT, novelty search, and Galactic Arms Race. His original 2002 paper on NEAT also received the 2017 ISAL Award for Outstanding Paper of the Decade 2002–2012 from the International Society for Artificial Life. He’s a coauthor of the popular science book Why Greatness Cannot Be Planned: The Myth of the Objective (Springer) and has spoken widely on its subject. He earned a BSE from the University of Pennsylvania in 1997 and a PhD in 2004 from the University of Texas at Austin.