Intelligent Robots

I was just reading a news article today which talked about a new type of soft robot. It would be able to walk not only over rough terrain, but through fire, ice, rain and almost any weather condition. It could even be run over by a car and still function. The article said this robot was being considered as a rescue robot. It sounds good to me, but one thing which seems to be lacking is the ability to pull someone out of danger. It didn’t look very big and from what I could tell which could be wrong, it seemed to be X shaped. I guess we will just have to wait on this one and see how things turn out.

Graphic Source: Ken, About Facts Org

A robot called OutRunner is being developed from Kickstarter funds and in the early phase, but it is an incredible machine. One of the reasons it is so incredible is the fact it can reach and maintain speeds of about 20 miles per hour. This is much faster than a human can sustain, so this means there will now be robots faster than us. We had a chance of getting away from a rogue robot before, because we were faster than robots, but those days seem to be gone. OutRunner is small, but the principal of its locomotion can be scaled up for larger robots.

When we speak about robots we shouldn’t forget a robot will only be as good as its brain. When we have a thought it activates thousands of brain cells which communicate with each other using electrical and chemical signals. Our neural network is the most complicated machine on the planet. Several years ago scientists began to think about creating artificial neural networks. There was a problem, no artificial neural network worked exactly the same way our human brains worked. Scientists worked on the problem and now they have come up with a new type of neural network called a spiking neural network. This network is much closer in operation to our own. Neurons in the artificial brain initiate signals to other neurons with a rapid electrical spike, like our brains do. This brain will be taught with a new algorithm which will allow it to decide what things it should learn and what to ignore.

The European Space Agency is developing robots to aid in disasters. The robots are actually controlled by a human wearing an exoskeleton. As the human makes gestures with the limbs on the suit, these are repeated by a robot. The Europeans believe this is the best way to use robots in emergencies, since there is no decision making on the part of the rescue robot. It will be much easier to send a robot to a repair job than a human. A good example of this is space telescope repairs. If this robot becomes operational there might not be any more need for an astronaut to repair space telescopes because the repairs can be accomplished by a robot directed by a human in the comfort of his ship or space station. Perhaps the human could even be on earth.

Scientists are now saying they have discovered consciousness is a state of matter and not a function of the brain which we all thought it was. If they are right we have a lot to think about. Yes, we could then say there is surely life after death, but we could also say maybe we could bring consciousness to machines once we find out how it works. If we brought consciousness to robots would this make them life forms? It would seem so wouldn’t it? Would we would then be bound by the same laws which govern us as to how we treat people? Think about this, a human gets tired of his or her robot and puts a bullet into its artificial brain, but the robot had achieved consciousness. Was this murder? That will have to be something for the courts to hash out.

Computers and robots are inexorably tied together. There is no doubt robots will have access to computers and may even have one built into them, even if they have a neural network. This will make robots much smarter than we are unless we all get chipped. IBM has announced their computer named Watson can now debate on any topic. Remember Watson won the Jeopardy challenge and beat Ken Jennings, one of the best Jeopardy champs of all time.
The rise of the robots has a lot of people running scared, even some respected scientists believe we will not be able to control them if they get too smart. Terminator brought this home to many of us and many who saw the movie and thought about it seemed to think it was only a matter of time before intelligent robots would decide we were in their way. This is one of those scenarios which has a feeling of realism to it. Will there be a way we will be able to protect ourselves from a robot catastrophe, or is it inevitable? I think this question would make for a great debate.

One of the problems with creating intelligent machines is we may never know if one becomes sentient. As computer programming improves, it becomes more impossible to tell the programming from the machines actual thoughts if it has any. In a very crude way this reminds me of the first programs on the Internet which you could converse with. You would type a simple question and either get a question back or an answer. It was almost like talking to a human if you kept things very simple. If this was done in a much more sophisticated way how would we know whether we were talking to man or machine, or even a sentient machine? The answer is we couldn’t be sure. Another problem as all we computer users know is things go wrong. What happens if we program an intelligent robot with ethics and tell in never to kill and something happens and its program gets corrupted and the never to kill instruction becomes kill, or someone hacks our robot and changes its ethics?

Robots are coming and in some cases are here already, but the ones around today are nothing compared to what is waiting in the wings.