Deceiving Robots
These intelligent machines might be one step closer to total domination thanks to a team of researchers at the Georgia Institute of Technology. The engineers designed what they are calling the world's first deceptive robot—one that can create a false trail and hide to avoid being caught.
"We have developed algorithms that allow a robot to determine whether it should deceive a human or other intelligent machine, and we have designed techniques that help the robot select the best deceptive strategy to reduce its chance of being discovered," said Ronald Arkin, a professor in the Georgia Tech School of Interactive Computing, according to the school's press release.
The research analyzed robot deception from a general perspective—robots deceiving people and robots deceiving other robots—in a step toward broad applications. For example, deceptive robots could be useful for search and rescue operations, in which a panicking victim might refuse to cooperate. A deceptive robot could also be invaluable for military purposes, such as hiding from soldiers, misleading the enemy and keeping intelligence information safe.
"Most social robots will probably rarely use deception, but it's still an important tool in the robot's interactive arsenal because robots that recognize the need for deception have advantages in terms of outcome, compared to robots that do not recognize the need for deception," said the study's co-author, Alan Wagner, a research engineer at the Georgia Tech Research Institute.
In their research, the engineers focused on robots' beliefs, actions and communications when trying to deceive another robot.
First, they taught the robot how to determine which situations warranted deception. The researchers created algorithms based on game and interdependence theory to test the value of deception. In order for deception to be initiated, the algorithm had to determine both a conflict and a potential benefit from deception. Next, the engineers programmed the robots how to deceive.
In their study, the researchers ran 20 hide-and-seek games with two autonomous robots. The hiding robot was able to successfully deceive (by hiding from) the seeker robot about 75 percent of the time.
"The experimental results weren't perfect, but they demonstrated the learning and use of deception signals by real robots in a noisy environment," said Wagner. "The results were also a preliminary indication that the techniques and algorithms described in the paper could be used to successfully produce deceptive behavior in a robot."
Although the benefits of deceptive robots will likely outweigh any negative consequences, the researchers do emphasize some ethical concerns.
"We have been concerned from the very beginning with the ethical implications related to the creation of robots capable of deception, and we understand that there are beneficial and deleterious aspects," explained Arkin. "We strongly encourage discussion about the appropriateness of deceptive robots to determine what, if any, regulations or guidelines should constrain the development of these systems."
"We have developed algorithms that allow a robot to determine whether it should deceive a human or other intelligent machine, and we have designed techniques that help the robot select the best deceptive strategy to reduce its chance of being discovered," said Ronald Arkin, a professor in the Georgia Tech School of Interactive Computing, according to the school's press release.
The research analyzed robot deception from a general perspective—robots deceiving people and robots deceiving other robots—in a step toward broad applications. For example, deceptive robots could be useful for search and rescue operations, in which a panicking victim might refuse to cooperate. A deceptive robot could also be invaluable for military purposes, such as hiding from soldiers, misleading the enemy and keeping intelligence information safe.
"Most social robots will probably rarely use deception, but it's still an important tool in the robot's interactive arsenal because robots that recognize the need for deception have advantages in terms of outcome, compared to robots that do not recognize the need for deception," said the study's co-author, Alan Wagner, a research engineer at the Georgia Tech Research Institute.
In their research, the engineers focused on robots' beliefs, actions and communications when trying to deceive another robot.
First, they taught the robot how to determine which situations warranted deception. The researchers created algorithms based on game and interdependence theory to test the value of deception. In order for deception to be initiated, the algorithm had to determine both a conflict and a potential benefit from deception. Next, the engineers programmed the robots how to deceive.
In their study, the researchers ran 20 hide-and-seek games with two autonomous robots. The hiding robot was able to successfully deceive (by hiding from) the seeker robot about 75 percent of the time.
"The experimental results weren't perfect, but they demonstrated the learning and use of deception signals by real robots in a noisy environment," said Wagner. "The results were also a preliminary indication that the techniques and algorithms described in the paper could be used to successfully produce deceptive behavior in a robot."
Although the benefits of deceptive robots will likely outweigh any negative consequences, the researchers do emphasize some ethical concerns.
"We have been concerned from the very beginning with the ethical implications related to the creation of robots capable of deception, and we understand that there are beneficial and deleterious aspects," explained Arkin. "We strongly encourage discussion about the appropriateness of deceptive robots to determine what, if any, regulations or guidelines should constrain the development of these systems."
Comments
Definitely gonna recommend this post to a few friends