Monday, March 21, 2011

Blinky, the bad robot

Combine HAL with R2D2 and you'd get something like Blinky, the star in the fantastic short film 'Bad Robot' by Ruairi Robinson. A must see for everybody on this planet, the film touches upon the soon-to-be unavoidable issue of machine morality, by showing how the behavior of a domestic robot takes an unexpected turn.

Sure we can just treat it as harmless entertainment, a purely fictional film for us to viscerally enjoy. But the more interesting and complete way is to reflect a little bit. So what can we learn here?

- Robots need social intelligence to be able to infer true intentions from people as opposed to irrational utterings that people do without any reflective connection. Probably sensing how calm people are will help robots do this. Another interesting ability for robots would be to detect whether or not somebody is in a state of close-mindedness or open-mindedness. For example, when a robot is constantly smiling, it will probably evoke negativity from a person in a close-minded state, but it can make a person in an open-minded mindstate turn more positive and search for inner ways towards happiness again. The detection of this mindstate is a form of emotional anticipation that can help the robot be a successful social actor.

- Many people will see robots as tools, and not as 'others' or 'equals', and treat them accordingly. They can often be judgemental, irritable, and full of expectations towards them, almost treating them as the ultimate promise of happiness. People who have that meme in their unconscious will, as the Dutch say, 'come home from a cold funfair'. It will take a while before people learn to see robots as they are, and develop a sense of true empathy and unconditional love for them.

- For robots, I believe that to a large extent true empathy and love can be programmed in the end. What is needed is a detailed understanding of the mechanics of consciousness, which will borrow from phenomenological data of human experiences, and will have to do with concepts from quantum physics, such as treating perceptional patterns as vibrations in order to understand them not as patterns of matter, but patterns of mind, that can give rise to ideas like matter. But let me not expand on this now. Basically robots will need to develop their social role from 'dumb servant' to 'loving and understanding father', and do this through learning about mindstates of people and what certain reactions will do to those mindstates.

If you have any reflections or comments, please leave them here so we can have some shared discussion into this fascinating topic. And maybe build that enlightened robot one day.

No comments: