IDG member Georgios N. Yannakakis just attended the Genetic and Evolutionary Computation Conference in Colorado, presenting original work from the Institute of Digital Games on evolutionary divergent search. The paper presented by Georgios is titled Surprise Search: Beyond Objectives and Novelty and introduces a new algorithm for Artificial Intelligence, which attempts to optimize the unexpectedness of the agents' behavior from one generation to the next. This work, funded by the AutoGameDesign project, has been carried out by Daniele Gravina and tests Surprise Search in a maze navigation task. The robots must learn to navigate a maze with dead-ends and deceptive pathways, with the goal of producing behaviors not expected based on their recent historical trends. Find more about surprise search here!