%0 Report %1 cs-97-197 %A Funes, Pablo %A Sklar, Elizabeth %A Juille, Hugues %A Pollack, Jordan %C 415 South St., Waltham MA 02254 USA %D 1997 %K Tron, adaptive agents, algorithms, autonomous coevolution, evolution evolutionary game genetic interactive learning, programming, robotics, software, %N CS-97-197 %T The Internet as a Virtual Ecology: Coevolutionary Arms Races Between Human and Artificial Populations %U http://www.demo.cs.brandeis.edu/papers/long.html#cs-97-197 %X we propose that learning complex behaviours can be achieved in a coevolutionary environment where one population consists of the human users of an interactive adaptive software tool and the öpposing" population is artificial, generated by a coevolutionary learning engine. We take advantage of the Internet, a connected community where people and software coexist. A new kind of adaptive agent can exploit its interactions with thousands of users-inside a virtual "niche"-to learn in a coevolutionary human-robot arms race. Our model is Tron, a simple dynamic game where introspective self-play quickly leads to collusive stagnation. We describe an application where thousands of small programs are sent to play with people through the Java interpreter running in their web browsers. The feedback provided by these agents is collected in our server and used to augment an ever improving fitness landscape for local robot-robot games. Speciation and fitness sharing provide diversity to challenge humans with a variety of differ ent strategies. In this way, we obtain an evolving environment where human as well as artificial adaptation are simultaneously taking place.

@techreport{cs-97-197, abstract = {we propose that learning complex behaviours can be achieved in a coevolutionary environment where one population consists of the human users of an interactive adaptive software tool and the {"}opposing{"} population is artificial, generated by a coevolutionary learning engine. We take advantage of the Internet, a connected community where people and software coexist. A new kind of adaptive agent can exploit its interactions with thousands of users-inside a virtual {"}niche{"}-to learn in a coevolutionary human-robot arms race. Our model is Tron, a simple dynamic game where introspective self-play quickly leads to collusive stagnation. We describe an application where thousands of small programs are sent to play with people through the Java interpreter running in their web browsers. The feedback provided by these agents is collected in our server and used to augment an ever improving fitness landscape for local robot-robot games. Speciation and fitness sharing provide diversity to challenge humans with a variety of differ ent strategies. In this way, we obtain an evolving environment where human as well as artificial adaptation are simultaneously taking place.}, added-at = {2008-06-19T17:35:00.000+0200}, address = {415 South St., Waltham MA 02254 USA}, author = {Funes, Pablo and Sklar, Elizabeth and Juille, Hugues and Pollack, Jordan}, biburl = {https://www.bibsonomy.org/bibtex/2aaabebc3f276fea544705540d7b9d079/brazovayeye}, institution = {Computer Science, Brandeis University}, interhash = {cbc878afadde6d2c147f1fd2001e1c5c}, intrahash = {aaabebc3f276fea544705540d7b9d079}, keywords = {Tron, adaptive agents, algorithms, autonomous coevolution, evolution evolutionary game genetic interactive learning, programming, robotics, software,}, notes = {See also \cite{funes_sab98} and http://helen.cs-i.brandeis.edu/tron/html/about.html Two populations: computer play (1000), playing against people (100). Generation based. non-standard selection and migration strategies. Deterministic play. Limited knowledge of game arena. Java. Problem with {"}live and let live{"} or conclusion between (evolved) players. p10 People played 22494 games in two months. p11 Marginal improvement in computer players (28% -> 35%). Humans better than computer.}, number = {CS-97-197}, size = {20 pages}, timestamp = {2008-06-19T17:39:56.000+0200}, title = {The Internet as a Virtual Ecology: Coevolutionary Arms Races Between Human and Artificial Populations}, type = {Technical Report}, url = {http://www.demo.cs.brandeis.edu/papers/long.html#cs-97-197}, year = 1997 }