It’s been the theme of many science fiction books and movies: humans build machines which are more powerful and ‘intelligent’ than humans. The machines learn to think and reason, and then assess that humans should be eliminated as a lesser life-form. This is the stuff of fiction now, but to some the fully cognizant machine isn’t that far away and what is in place to protect humankind?
Scientists at the University of Hertfordshire, which is based in Hatfield in the U.K. have proposed an answer. This is with the development if a concept called Empowerment. The aim is to use this to enable robots to protect and serve humans, while, at the same time, keeping themselves safe. The aim is to enable thinking machines to continually seek to keep their options open, and doing the same for the humans around them.
This dilemma is of significance given the rapid technological pace with the development of robotics. What, for instance, should happen with autonomous vehicles? If the choice is between crashing into a wall (which could kill the human occupant) or into a cyclist (which could kill a different human but protect the passenger in the self-driving car), what should the machine do? Alternatively, is the object of the machine powering the self-driving car to keep the car as safe from damage as possible or to protect humans who might be injured from the vehicle?
READ MORE: Is it possible for a robot to write the news?
According to researcher Dr. Daniel Polani, from Hertfordshire’s university, who told Now Science: “Public opinion seems to swing between enthusiasm for progress and downplaying any risks, to outright fear.” To come up with something to reassure the public the researcher went back to Isaac Asimov’s three laws of robotics.
Asimov’s three laws of robotics:
A robot may not injure a human being or, through inaction, allow a human being to come to harm.
A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
A robot must protect its own existence as long as such protection does not conflict with the First or Second Laws
From the “Handbook of Robotics, 56th Edition, 2058 A.D.”
To get close to this, Dr. Polani notes is that the concept of “harm” is complex, context-specific and is difficult to explain, on the current basis of machine intelligence, to a robot. The consequences are that if a robot does not understand “harm,” how can the robot then avoid causing it?
READ MORE: People prefer slightly faulty robots to perfect ones
The researchers have devised a program and logic system called Empowerment. This takes the approach that instead of attention make a machine understand complex ethical questions it is better to crate robots that seek to keep their options open. This means a state where the machine has the greatest potential influence on the world it can perceive. What is needed, the computer scientist argues, is for the thinking robot to also seeks to maintain a human’s ‘empowerment.’ “This means also seeing the world through the eyes of the human with which it interacts. This is an altruistic concept and, if taken up, ‘Empowerment’ might form an important part of the overall ethical behavior of robots.
The concept has been published in the journal Frontiers in Robotics and AI, in a research paper titled “Empowerment As Replacement for the Three Laws of Robotics.”
