[Tyler]

Teaching Robots to Lie
1 April 2013, 7:23 pm

 

Learning to lie and deceive others is as basic to human social interaction as cooperation. Deceiving opponents and detecting possible deception is basic to all sports, for example. Lying through omission in daily conversation in order to avoid hurting someone's feeling is another common example. Researchers refer to these as "deceptive and non-cooperative behaviors" and robots will need to learn to deceive and detect deception if they're going to keep up with humans in social situations. The IEEE Intelligent Systems journal devoted an issue to this topic in December of 2012. The journal is normally pay-walled but the Computational Deception and Noncooperation issue has been made open access through the Computing|Now website and it's well worth a read. It includes article on modeling and detecting deception, biologically inspired robot deception, deception in sports and immersive environments, non-cooperative agents, non-verbal clues to deception, and a few other interesting topics. You can bet this issue is on Bender's reading list._{CC BY 2.0 photo of Bender by Liam Daly}

Source: robots.net

To visit any links mentioned please view the original article, the link is at the top of this post.