illuminating science

I’m a big sci-fi fan, at least of good sci-fi with a strong physics background. Today, in a discussion on The Dividing Line, we were talking about whether telling someone the truth is always the best option, or (sort of) more quantitatively “does the least harm”. This inspired me to do a little digging on Asimov’s Three Laws of Robotics, and I found this fascinating piece from about 10 years ago on how relevant these laws are to modern robotics. For the record, the Three Laws are:

1. A robot may not harm a human being, or, through inaction, allow a human being to come to harm.
2. A robot must obey the orders given to it by human beings, except where such orders would conflict with the First Law.
3. A robot must protect its own existence, as long as such protection does not conflict with the First or Second Law.

Many of Asimov’s plotlines came out of exploiting apparent paradoxes and contradictions from within these Three Laws. For instance, if I were about to be killed by a maniac with an axe, a robot might shoot the maniac in the leg, even though it causes him harm, in the hopes of preventing greater harm to me. In Asimov’s world, the laws are hardwired in the robot’s brain and “positronic potentials” tip the balance towards one act or another according to those laws. Many times, faced with an impossible decision, the robot simply fries its circuitry. The article is a really interesting summary of some of the issues with the laws.

In the article, Roger Clark discusses how the Laws of Robotics have found their way into the thinking of many modern roboticists. He also suggests some interesting “meta” laws that might be needed to constrain the robot in our world, where we don’t have positronic brains available! And there’s a few interesting points on the human acceptence of robots, where he suggests that we need to educate people of the limitations of robots, and computers too, so that people don’t over identify with them. Although written in 1994, this is perhaps becoming increasingly relevant with robot dog pets becoming popular. And that makes for an interesting question - if a child becomes attached to and loves a robot dog, is that any worse than a real dog? Or a teddy bear? Asimov has a story where a young girl becomes attached to Robby the Robot, much to her parent’s dismay. In the end, she loves the robot as much as any pet, and they realise that Robby does have a place in the family.