Researches in Switzerland have found that robots equiped with artificial neural networks can develop the ability to decieve under guidence of simple regulations.
The robots are programed to seek out food and they learned to conceal their visual signals from other robots to keep the food for themselves. The team made generations of these robots by combining the AI neural networks of the most successful robots. A few random changes in code models the idea of biological mutations. Interestingly, after 100 generations the robots stopped flashing light when they are near food and evolved to attract or repelled by light. Due to the competition for food, the robots try to conceal information that gives it away.
This research can further our understanding of biological communication systems.
Reference: Kristina Grifantini ~ Technology Review.com
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment