Robots can learn to lie and deceive
Researchers at the Federal Polytechnic School of Lausanne (Switzerland) have developed robots that resemble people. However, the similarity is not in external signs, or, for example, the ability to think or experience emotions, but the ability learn to lie and deceive. Miniature robots on wheels are programmed to search for food (of course, not present). In addition, they are equipped with alarm LEDs are flashing blue when it detects food. Cameras other robots can fix the flash, and close to the same place, it allows them to find food more quickly. As a result, some “feeder” inevitably formed collision.
Scientists have introduced robots in an artificial neural system able to grow and learn. Of course, to the level of rights to its very far, however, understand that the outbreak of blue increase competition and reduce the chances to get food, robots could. Developers copied the code’s most successful robots with multiple random additions, simulating the evolution and the “output” new generation of robots. As a result, they have learned to hide the location of the “trough”, not indicating this to other robots. After several hundreds of “generations” Most robots are no longer included the blue indicators. Why scientists robots that can lie, remains a mystery. Perhaps the way they had planned to approach the creation of artificial intelligence similar to human.
Source
Researchers create robots that can learn to lie and deceive, on purpose
Share Your Opinions about “Robots can learn to lie and deceive”