Artificial intelligence is commonly abbreviated as AI. It is a discipline that has its roots in philosophy because philosophers for over 200 years have been studying the issues to do with sense, learn, thinking and remembering mechanisms. The theory of learning and reasoning has been under study for 200 years where several theories have been formulated. Some of the theories concluded that reasoning is made of several physical systems. Through such efforts there has been several theories developed such as the theory of probability, logic, decision making and calculations from the field of mathematics. The development of first generational computers brought to live the concept of scientific analysis that relates to intelligence by putting the theory into practical in 1950’s. It was the belief of many scientists that computers will offer practical knowledge on electronic thinking with a super speed. However, electronic intelligence proved to be very difficult to implement those early days (Luger, 2005). This is because of the processing speed and storage devices of the first generation computers.
AI is the concept where electronics device mind sense, understands, predict and manipulate concepts in the world that are larger and more complex than itself. The name Artificial Intelligence was formerly coined in 1956 making it the newest scientific field of study. The developments of AI can be categorized intro three periods (Provan, 1992).
The first period or generation of AI was during its inception through to 1960s. This was a period where there was much enthusiasm and great expectations. This is because it was a new field just formulated and there were a lot of expectations from the stakeholders. AI was very successful during this period because of the limited capability of the computers. Computers in this period were considered as primitive and only capable of carrying out arithmetic operations.
Some of the AI systems that were created during this time where the world champion chess software and the universal translator software. Allen Newell and Herbert Simon created software that was the first one to imitate human protocol in solving general applications. This AI program was referred to as General Program Solver (GPS). The systems were able to prove mathematical theorems such as the Gelernter’s Geometry Theorem Prover. During this period Arthur Samuel also developed a program that played Draughts, the program matched with the high levels of its competitors. This program has become a better after playing with it and was able to play better that its creator Samuel. This gave the idea that AI systems do not only do what they are told to do but they can also learn from experience and improve. John McCarthy created AI programming language in 1958 which become the primary AI programming language known as lisp (Russell et al., 1995).
The second generation of AI development was from mid 1960s to 1980s. This period can be described as the period of disillusionment and development of knowledge based systems. The first generation AI system that was designed for general purpose was only successful in performing simple tasks but failed to perform when it came to complex tasks. The major reason for failing was that early software programs were very simple and could not take in complex inputs. The early programs also were made to compute simple step sequence and they were experimented based on the simple sequences hence they failed on complex cases.
The expert systems were realized at the end of 1960s. The systems were developed based on rule-based knowledge base where they had to reference the knowledge base in their decisions. This resulted in the development of systems that would solve the uncertainty because techniques for mapping of knowledge representations and heuristic search methods were developed. MYCIN system is conserved as the first expert system which had 450 rules designed on order to match up to the human experts. Logic programming language was called Prolog was developed in the 1970s and facilitated the inclusion of calculus in computer programs. Prolog is being used in the development of expert systems in medical field and other disciplines. The second period had improvement in the AI field after the development of Prolog, expert systems and the natural language parsers.
The third period is where AI is realized into a full industry of its own. This is after further development of the expert systems into R1. The AI industry became more profitable where 40 million dollars was the yearly savings for a development company called DEC. there were about 40 expert systems developed and deployed to different industries by the year 1988 by the DEC’s AI group. The fourth generation computers were announced by Japan in 1981. This project was used to develop an intelligent computer that would answer to Japanese language. This prompted other countries, the USA and other European countries develop intelligent systems with similar scope as the Japanese project. The projects were designed by Prolog programming language.
Through this project resulted into AI being taken out of the laboratory into different fields such as medical, judicial, robotics, geology, chemistry, industrial process control among others. The industry of AI grew and there were a lot of application of the system in different industries. The annual income grew to 2 billion dollars in 1988. There was another discovery during this period. The statistical AI-methods and the discovery of neural networks boosted AI development.
AI technologies have expanded their scopes from a single field to everyday life. AI is being used in speech recognition, handwriting-recognition among other areas. This is possible because of the early discoveries of the hidden Markov-models. The field of robotics has been revolutionized by many discoveries. The field of robotics, machine vision and learning are what were are experiencing now. Currently high speed processors and hardware devices are being used to develop very powerful AI systems. The graphical user interface design has also enabled quick use of AI tools in many fields. Unlike the early software systems , the current ones can be re-engineered and upgraded hence adding new functionalities.
The future of AI in our society is beyond imagination. There is going to be an improved relationship between machines and humans where machines will not only work as human aids but they are going to replace human activities. Humans are going to speak to the computer and instruct it what to do. Robots are going to have human intelligence and all of the human senses. The diagram below shows the relationship between IA and other fields.
Fig 1 AI and its application
Li, D, Liu, CY, Du, Y, & Han, X 2004, Artificial Intelligence with Uncertainty, In CIT (p. 2).
Luger, GF 2005, Artificial intelligence: Structures and strategies for complex problem solving, Pearson education, New York.
Provan, GM 1992, The validity of Dempster-Shafer belief functions, International Journal of Approximate Reasoning, 6(3), 389-399.
Russell, SJ, Norvig, P, Canny, JF, Malik, JM., & Edwards, DD 1995,.Artificial intelligence: a modern approach (Vol. 74),s Englewood Cliffs: Prentice hall.
Van den Besselaar, P, & Leydesdorff, L 1996, Mapping change in scientific specialties: A scientometric reconstruction of the development of artificial intelligence. Journal of the American Society for Information Science, 47(6), 415-436.