Asimov, the well known science fiction writer long ago framed that man will one day be able to build robots which are as intelligent as or much more intelligent than humans. He also foresaw the need to maintain the superiority of the humans amongst the robots and to make sure that the robots never take over the humans with their advanced intelligence mechanism.
Obviously, humans at any cost want our supremacy to continue on all our creations and we wish to see the robots as our slaves always and not the other way round. If you have seen the movie, Terminator , you will have an idea as to how dangerous it could be if robots take over human intelligence.
For this Asimov suggested a set of three basic laws to be fed into the basic root of any robot action, which any robot can never violate. Any robot before committing any action has to first ensure that the action does not violate any of these basic laws.
The three laws of Asimov are as follows:
1. A Robot shall never carry out any action which will cause any injury to any human being.
2. A Robot shall always obey a human being’s orders provided the first law is not violated.
3. A Robot shall always try to protect itself, provided the first two laws are not violated.
It is quite obvious that these laws are intelligently framed to avoid any robot intelligence from taking over human intelligence and trying to rule or ruin the human species. The priority of these laws for the robots would be in the descending order i.e. first law has a greater priority over the second and the second a greater priority over the third. The third law has the lest priority among the three laws.
The first law clearly bars any robot from causing any injury to any human, there by making it impossible for any robot to even think about any act which would be harmful to any human, by eliminating in the initial stage itself. Thus, all those robots whose intelligent circuits or software brains are based on this law will never even think of causing any harm to a human.
Then comes the second law. The intention of this law is to ensure that robots are never allowed to become superior to humans. They shall continue to be the slaves of humans, no matter how intelligent they become. A robot is always supposed to carry out any order directed by a human. At the same time in this law, it is also taken care that humans do not use robots for their inter-human rivalry where in one human orders a robot to injure or kill another human being. So the second law clearly states that the robot shall always follow the directions of a human being, provided he does not direct the robot to injure or kill another human being. In other words, the second law is valid provided the first law is not violated.
It should be noted here that, humans can direct a robot to injure, destroy or kill another robot and as there is no law which bans this act, the robot shall be compelled to carry out such an order by humans. Humans can issue such an order in case there arises a situation where in some rogue robots harmful to humans get created by human fault or intentional sabotage by some humans.
Then there is the third law, which now talks about the protection of robots themselves. Certainly, these intelligent robots being one of the most valuable creations of human intelligence, these robots have to protect themselves and survive. Hence, whenever there is a danger to the life of a robot, it has all the rights to protect itself, provided the first two laws are not violated.
A robot cannot protect itself when there is question of either the robot or a human will be saved. In such a situation, the life of a human becomes more important for the robot than its own life. Thus the robot shall not follow the third law if the first law gets violated.
Also, if a human asks a robot to destroy itself, then the robot has no right to protect itself as the second law again has a higher priority over the third law and the robot has to carry on the human’s order. This will be of particular use when the destruction of a robot would save human life or valuable property. For example: a robot may go mad if some of it’s circuits get short and may prove costly to human lives around there. In such a case a human might just order the robot to destroy itself by preventing the robot from causing any damage to the humans or property nearby.
Although, these laws may seem to be quite sufficient to maintain the human supremacy over the robots, there might arise a situation where a power hungry human wants to take over the either a part of or the entire world and destroy all or a part of human population. Hence, later Asimov introduce one more law called the zeroth law of asimov which has a higher over all the other three laws. The zeroth law of Asimov states that, A robot has to always ensure the survival of the human race.
Now the complete set of Asimov’s laws are follows:
0.A robot has to always ensure the survival of the human race
1.A Robot shall never carry out any action which will cause any injury to any human being, provided the zeroth law is not violated.
2.A Robot shall always obey a human being’s orders, provided the zeroth and the first laws are not violated.
3.A Robot shall always try to protect itself, provided the first two laws are not violated, provided the zeroth, first and the second laws are not violated.
In case there is a evil human trying to eliminate a part of or the entire humanity, the robot is entitled to carry out an offensive action against such human based on the zeroth law of Asimov. But then as one can again see, it is quite easy to mislead a robot into believeing something and then use it as a weapon to achieve one’s own personal goals!!
So the real challenge will be the coding of the Robot’s analysis software which has to be intelligent enough to verify things without being mislead, which definitely is a tough goal to achieve!