Artificial Intelligence

Intelligent Machines

Soumyabrata PaulSoumyabrata Paul

Humans are the most intelligent among all the living beings in the world. At least we, the humans think so. From the initiation of our civilization, we have been developing numerous instruments. These instruments were used in daily life and as well as for the servival. We never stoped and kept dveloping the old ones and inventing new ones. With the development of instruments we build devices. And complicated devices formed machines. We had been trying to develop such system that not only be used to accomplishe tasks but also help us to do much more complicated job for human. Thus machines evolved. Still we are not satisfied. We are now trying mak to make the machines as intelligent as ours. We cannot feed organic brains to the machines. But still we are on the way of make them intelligent with logical reasoning, mathematical calculation, sensing environment with various input devices and various complicaed algorithms. The intelligence, thus building up in the machines is known as Artificial Intelligence.

Machines are with us from eras. But things started changing with the advent of Electronic Computer in 1941. A huge information and related program can be stored in a considerably small volume with Stored Program Computer, developed in the year 1949. Norbert Wiene, an American scientist, was working with the Feedback Theory and theorized that every intelligent behaviour is a result of feedback. Let us have an example, say BiMetal-Strip used in an machine, the activity of which depends upon the temparature of its surroundings. The machine acts according to the status of the BiMetal-Strip and BiMetal-Strip gives feedback about the temparature. John McCarthy is known as the father of Artificial Intelligence. The term "Artificial Intelligence" was first coined by him in "The Dartmouth Summer Research Project on Artificial Intelligence" at Vermont, in the year 1956.

In our present technological advancement, we have developed the AI (Artificial Intelligent) of machines to significant level. But still there are much more to be developed. We want to make the machine intelligent like human. There is neither any means nor any scale according to which we can measure the level of intelligence. Yet we try to follow a relative measurement known as Intelligent Quotient (IQ). The major drawbacks in machines with AI are "learning from experience", "inference" and "common sense knowledge and reasoning".

In our childhood, we can simply be compared to simple machines. But as we grow we develop our intelligence. How do we develop our selves without supervision ? Here comes the "learning from experience". Every monent we store our experience from an incident and result into our memory. But in case of machines we need to feed situations and corresponding descisions as program logic. With the growing memorized experience, we do guess of an unseen object. This is known as "inference". When we hear the word "bird", we reliase an living object that is covered with feathers and can fly. But in case of Machines if all the informations regadirding all the species of birds had not been feed it will not the type of object "bird". If this is fulfilled and furthermore we mention the name of the bird, the machine will identify the "exact object". Thus machines fail to do what we call "inference".

Another major drawback is "feeling". We can feel but the machine cannot. This is not what we feel with our sense organs. The machine can also sense with devices like camera and many other different sensor devices. What we are talking about is our inner sense, what we feel from the heart and mind. Let us say I am talking to an unknown person, I can feel whether I should make a friendship or simply avoid the person. This feeling has never been felt by any of intelligent machines. The most challanged drawback is "common sense knowledge and reasoning". The theaory interprets to follow non-monotonic reasoning. Like human, the machine should apply the most suitable common sense from a heap of defined common sense facts applying non-monotonic reasoning, not the Logic AI. Researches are going on to overcome all the drawbacks of intelligent machines. Even genetic engineering involves with AI when we try to implement intelligence with organic or semi-organic bodies.

Many of us believes that with the development of Robotics and Artificial Intelligence, one day the machines will become so much human like that they will become intolaurable to human, to the mankind. As a result, war around the world will be unavoidable. Even this theme has been in movies like "Terminator". But who cares. We have gone far in AI but still miles to go. In future surely hopes to see machines as our beloved friends.