Cal Tech has announced the “Turing Tournament,” designed to “find the best computer programs to mimic human behavior … and the best computer programs to detect the difference between machine and human behavior.”

Two types of submissions will be accepted: an emulator, which generates a dataset that mimics human behavior, and a detector, which detects the difference between datasets generated by human and machine behavior.

“The main difference between our Turing tournament and other Turing Tests, including the Loebner prize, is that we in fact do not have a human interrogator who is asked to tell the difference between machine and human,” said Jasmina Arifovic, Director, Turing Tournament Project.

“Instead, we ask people to write algorithms that we call ‘detectors’ that would look at machine and human datasets and try to tell the difference.” These play the role assigned to a human interrogator in the original Turing Test.

“We chose to implement the tournament to study learning in repeated games. The reason for this decision was that there has been a lot of research interest recently in modeling the behavior of agents in this type of environment. This means that there are already a number of ’emulators’ that people could start working with.

“The human behavior in our tournament is represented by datasets that will be generated in the laboratory experiments with human subjects. We intend to conduct these experiments (at Social Science Experimental Laboratory at Caltech) after the May 31 deadline. Human subjects will participate as players in a number of repeated games that will be the environments in which emulators and detectors will compete.”