Asimov himself proved that there are loopholes in his three (and later 4) laws of robotics.
In one of his stories a character used a robot to commit murder. He told the robot to put the powder from the bag on the top shelf of a certain cupboard into his wife's coffee, and then serve it to her. Robots would have to be way smarter than people to realize that a human would come to harm from that action.
The problem is that if a robot cannot tell right from wrong then the robot can not follow the three laws no matter how hard it tries.
Actually, IR is not the best method to sense the candle. There is a Hamamatsu sensor called a 'UV-Tron' that senses the UV light given off by the candle. Many of the higher end robots use the UV-tron to detect whiether the candle is in the room, just by looking in from the doorway with the UV-tron. It is not all that directional so for the final approach people do something else, IR, or look for the candle circle. My approach would be unique (when I get it working) that I only can find the candle stick itself.
I have been building robots for years and finally wanted to compete at Trinity, as I had read about Trinity every year. I designed my robot, Tryclops, to win the expert division. Unfortunatly I did not finish my sensor (a laser range finder that returns 64K ranges (1 cm resolution) per second). I went anyway (flew in from Minnesota) and showed off what I had and wandered around and talked to everyone. Some of the seminars were very interesting, All the robots were interesting. I plan on going back next year with a finished Tryclops
Slashdot Mirror
In one of his stories a character used a robot to commit murder. He told the robot to put the powder from the bag on the top shelf of a certain cupboard into his wife's coffee, and then serve it to her. Robots would have to be way smarter than people to realize that a human would come to harm from that action.
The problem is that if a robot cannot tell right from wrong then the robot can not follow the three laws no matter how hard it tries.
Actually, IR is not the best method to sense the candle. There is a Hamamatsu sensor called a 'UV-Tron' that senses the UV light given off by the candle. Many of the higher end robots use the UV-tron to detect whiether the candle is in the room, just by looking in from the doorway with the UV-tron. It is not all that directional so for the final approach people do something else, IR, or look for the candle circle. My approach would be unique (when I get it working) that I only can find the candle stick itself.
Quite a few people do just look for the candle circle. You'll notice that in the 'expert' class, there are no candle circles
I have been building robots for years and finally wanted to compete at Trinity, as I had read about Trinity every year.
I designed my robot, Tryclops, to win the expert division. Unfortunatly I did not finish my sensor (a laser range finder that returns 64K ranges (1 cm resolution) per second). I went anyway (flew in from Minnesota) and showed off what I had and wandered around and talked to everyone. Some of the seminars were very interesting, All the robots were interesting.
I plan on going back next year with a finished Tryclops