Artificial Intelligence (AI)
Artificial Intelligence is one of the hot topics of today’s IT world. The name itself suggests that it is intelligence which is artificially created.
As we all know for human beings intelligence could not be or need not be created artificially, as it is natural. So it is sure that artificial intelligence is for machines but we cannot be specific while defining intelligence. When we consider a human being, the parameters that are considered for defining his intelligence will vary from person to person, that is it is an immeasurable characteristic.
Still we could say that intelligence helps to act appropriately in appropriate time using the previous knowledge and experience. Technically speaking, Artificial Intelligence (AI) is the branch of computer science which focuses on creating intelligence to machines. This study started long back in 1956 by John McCarthy, but even today its importance is radically increasing.
The first image that comes to our mind when we hear the word Artificial Intelligence might be a robot. Yes we have seen movies in which robots work as human beings or even play table tennis. When we just think of implementing it, we might feel it as an easy process. But it is very complex to simulate machines like human beings in case of intelligence. We never think of the complexities involved or the process happening within our body, when we do even simple actions. If it is the case with a human being with natural intelligence, the steps will be too difficult when it comes to the case of machines. It involves a long set of procedures and algorithms.
To understand the techniques used to implement AI, we should have a clear idea about the way in which our brain works. A brain is formed using millions of cells, biologically termed as neurons. Many signals are passed through these neurons whenever some information needs to be entered into our body. Based on this information we react accordingly. So in case of Artificial Intelligence mainly two approaches are used to implement intelligence. They are bottom-up and top-down approaches. In bottom up approach exact model of the human brain is formed where as in top down approach using computer programs the brain behavior is copied.
The main factors that need to be considered while simulating intelligence are:
- Representing Knowledge
- Step by step reasoning
- Set Goals and achieve them
- Learning new techniques
Artificial Intelligence cannot be implemented using normally used programming languages. There are some languages really intended for this purpose. As the logic and problem to be implemented itself is vague, the programmer should be able to concentrate completely on the issue and he should be free from all other technical complications. So languages which support declarative programming style are preferred in this case. Lisp and Prolog are the commonly used programming languages for AI. Lisp is a pure functional programming language where as Prolog is a logical programming language. Lisp stands for List Processor and Prolog for Programming in Logic. Even some other languages are available which are not that common. They are IPL, STRIPS, and Planner etc. Also C++ and MATLAB can be used for the same purpose.
Now after 50 years of study in the field of AI, we could see machines which play games like chess which demands intelligence, computers which understand human language, machines helping to make decisions in real time situations, robots which reacts instantaneously depending on the input they receive etc. Even though we could not design machines which exhibit intelligence completely, most of the machines are really smart. There are machines which defeat intelligent human beings in chess games. Now robots are being used in plants where there is risk for human life. Designing computers which understand human language can be considered as the most complex aspect of Artificial Intelligence. Also this implementation would make robots very familiar to common people who don’t have any expertise. You might have come across systems which allow you to generate voice messages for the corresponding entered texts. Some other applications which introduced AI are computer design, training, weather forecast, nuclear monitoring, telephone repair etc.
There are number of advantages by using AI. Tasks which are extremely complex and needs high level of intelligence can be completed in the blink of an eye using AI implemented machines. It might demand hours of human word to complete the same task. Duplication is easier in case of machines. Rather than training a large number of employees for a particular task, a single AI machine can be trained to achieve the same result in an easier and effective manner. In decision making process, where the decisions depend on the human emotions, it is better to depend on machines as they think more logically than emotionally. For e.g., regarding an operation needed for a baby of 3 months, even though the success rate is 60% while considering the facts, doctor might be held back because of many emotions. But if the decision is to be made by a machine, it will surely go for the surgery considering the facts. Moreover as it is implemented on machines it can work continuously without any break and the productivity will be the same from the starting moment till the work is finished.
There are some disadvantages also for this. Firstly, as always in the case of machines, there is chance to damage and hence for loss of data. Thus, the entire effort goes in to vain. Moreover, it needs regular maintenance and it surely affects the productivity. The amount of sensing input that can be taken is less compared to humans. Or else it has to connect to individual input devices for each purpose which is not practical. Another emotional issue needs to be considered is that it gives humans a feeling of insecurity and inferiority as by the introduction of machines many might lose their job. By considering all these factors, if we reach at an appropriate decision regarding the areas in which we have to introduce AI, then it is one of the incredible techniques the world has ever seen.