A new scientific paper has come out which studies the ability of computers to become conscious and talks about the “Phi” theory. In case you have never head of this theory it refers to a theory which is a complex theory of consciousness. It asks the question, is a computer controlled machine which is programmed to do a certain task conscious? Is an insect which goes through the daily routine of gathering and even growing food as some ants do conscious? Followers of Jainism a very ancient religion believe every living thing is aware. This is one of the things we have been arguing over for centuries. The bottom line is no one can say for sure if insects are aware, but most people believe their machines are not. This is not to say there could be a time in the future when they do become aware. If a computer became aware what would it think about us? Would it have emotions, or even opinions about things? We may be closer to the answer to this question than we would like to be.

The paper espouses a new theory which states computers might become self-aware. If this happens then we will know this is not unique to humans and might also be possessed by animals. We have always thought only humans knew about death, but when Koko the gorilla, who can talk sign language, was asked about death the gorilla replied it was being buried in a warm hole. While this is not exactly the clearest of explanations, it does seem to indicate some understanding and if one studies gorillas they are known for burying dead birds and such. Perhaps they understand far more than we give them credit for. Does an elephant know what death is? Many an elephant has been seen crying over the death of a loved one. I am talking about actual tears streaming down their faces. So what about this new theory?

The theory states, as far as I can tell, the more data is stored the more chance of becoming aware, but it has to be more than even that. For example if one has a storage device and fills it with terabytes of data, it still will never become aware, but a complex mammalian brain could be. There have been quite a few scientists who believe dolphins are aware and it is true they do some very intelligent things. Many a human has been saved from a watery death by dolphins who have either brought them to land or fought off sharks for them. This hardly seems the work of an animal which isn’t aware. One of my favorite dolphin stories is about a dolphin and a dog. Every day at the same time a dolphin swims to a dock in a fishing town where the same dog is there to greet it. The dog jumps into the water and the two of them play for a while and the dolphin leaves, but always comes back the next day. The two animals couldn’t be closer friends. It is obvious both of them are looking forward to their encounters so couldn’t one assume they are aware?

Being able to tell if a computer becomes aware may be an impossible task. There are certain tests which can be administered, but it seems to me a clever programmer could fool the tests by making us think the computer is aware when it is really not. Take the IBM computer Watson for example. Talking to Watson is quite an experience, it certainly seems alive and aware and if one didn’t know they were talking to a computer they might believe a human was on the other end. Watson is definitely not aware, but is so cleverly constructed one might believe it is. Some might argue perhaps it is aware and we just don’t know it. Watson does learn as it goes along and so do children and at some point they become aware so it might be argued why can’t Watson. Watson is being called a cognitive assistant and scientists are saying with the help of Watson any scientist can be as good as the best one in the field. Much as a human does, Watson uses unstructured data and analyzes it. A human might read a newspaper or an article on the internet and draw inferences from it. Watson will read articles and also draw conclusions from what it reads, only it will do it far faster than any human could. It will analyze a complex question to make sure it understands the meaning, much as we do and then check its solutions and answers for the best fit.

One of the problems we face with machine intelligence is when the time comes and a machine finally becomes aware we have to give it a conscience. Without one it will think it is free to do anything it wants. There is also the law of robotics put forth by Isaac Asamov which should be incorporated into every robots operating system and it is:
· A robot may not injure a human being or, through inaction, allow a human being to come to harm.
· A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
· A robot must protect its own existence as long as such protection does not conflict with the First or Second Laws.

This is where one of the problems comes in. We are already using robots to kill people. The military has all sorts of robotic machines which are programmed to kill autonomously. If machines learn as we are now being told, could we be teaching them the wrong lessons and sowing the seeds of our own destruction? Think of it, if it is okay to kill some humans why would it be wrong to kill others, remember we are dealing with machine logic here?
If it is true computers will become aware someday, we have to be very careful what we teach them.