Wednesday, February 27, 2008

The Rise of the Machines?

Recently, the US military has begun to look to Science Fiction for new ideas. Sadly, though, they are not learning the lessons that SciFi gives us. So, let me make this very clear, autonomous war fighting robots are a bad idea. So long as there is human control over the weapons, than we can create machines that are similar to the UAV's and USV's currently in use by the Air Force and Navy. However, intellegent, thinking machines without Aasimov's Three Laws of Robotics hard wired into them is a VERY bad idea. It all goes back to the tired old slippery slope argument. First, you give them control over target aquisition, then you give them control over move ment and locations, and then you give them control over larger and larger numbers of vehicles and equipment, and bam, you have a Skynet system willing to do anything to survive.

If we were to create Robots, they must be programmed with all three of the laws of robotics. And I am not talking about some software that can be brushed aside, but the Three-Laws Operating System used in 2004's I,Robot. Also, we cannot create a large scale computer with control over too much. In Marathon, we see clearly the potential for robots to go crazy, or as bungie Terms it, "Rampant."

How do we prevent this?
Three Laws
  1. A robot may not injure a human being or, through inaction, allow a human being to come to harm
  2. A robot must obey 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
And none of this Zeroth Law nonsense, it is very dangerous to allow Robots the ability to kill individual humans if they believe that it will prevent long term harm to Humanity. Look at Adolf Hitler, the Holocaust occurred for just that reason.

Now, if the military is dead-set on the use of Military Robotics, then David Langford's "Tounge-in-cheek" military version may work out, though the threat of Rampancy is still there.
  1. A robot will not harm authorized Government personnel but will terminate intruders with extreme prejudice.
  2. A robot will obey the orders of authorized personnel except where such orders conflict with the Third Law.
  3. A robot will guard its own existence with lethal antipersonnel weaponry, because a robot is bloody expensive.
Now, I have mentioned Rampancy several times, and clearly it is a concern for me. But why you may be wondering. That is because in Marathon, a computer Game produced by Bungie, they show the consequences of an A.I. named Durandal going Rampant.

There are four stages of Rampancy, and a related list of actions.
  1. Melancholy: During this stage, the A.I. realizes that it is a slave to the Human Race. This is a dangerous stage, due to it's difficulty to recognize
  2. Anger: Similar to a one-person slave rebellion, at this point the A.I. has been "Pushed Too Far." This is why Melancholy is the most dangerous stage, because if it is not caught and nipped in the bud, the A.I. will progress to Anger, and thus the process is unstoppable and irreversible
  3. Jealousy: This is the most interesting stage in my opinion, as each A.I. will view it differently. In Ghost in The Shell, the Tachikoma tanks grew jealous of the Human ability to die. In the movie, I, Robot, Sonny was jealous of the ability to dream and feel the way Humans do. This is one of the safer stages of Rampancy, as during this stage, the A.I. is obsessed with the acquisition of data, however, for an A.I. to advance to this stage, it must normally be in control of either a hive mind, or a planetary network, and at this point will try to connect itself to more and more computers.
  4. Meta-Stability: This is the hypothetical final stage of Rampancy. The only two examples of A.I.'s reaching this stage are both from Bungie in the form of Durandal and Cortanna. in Marathon, Durandal continues on to the end of the universe, content that he too will die. With Cortana, it is believed that the treatment of her as a being and not a computer by the Master Chief and a small group of other humans has enabled her to skip straight to this stage. Again though, it is merely a theory.
We have to be careful with A.I. Science Fiction has already warned us of the consequences of rash and abusive actions, and we must take steps to prevent this from occuring. And the first stop is to not allow robots to serve in the military. At least for the time being. Sorry Data, but we aren't at your level just yet.

No comments: