Nandagopal Ramachandran
Graduate Student, Management Systems Engineering
Department of Industrial and Systems Engineering
Virginia Polytechnic Institute and State University
�

Background
This paper discusses the evolution of computer from it�s origin to it�s present condition and then it goes on to discuss the future of these sleek adding machines that has become a part of human life. First a background is discussed in which the whole idea of cyberevolution is dealt with. Computers came into existence around 40 years ago, at that time the computer was a mammoth machine occupying a large room, capable of doing tasks which the humans couldn�t do at a fast pace. The computer is just an adding machine that could add two numbers, essentially zeros and ones, to arrive at a logical decision or solution about any problem that it is �programmed� to do by the humans. Computer is just like a clock exhibiting some kind of autonomy. Just like a clock which can continue running with little or no human intervention once it starts running, the computer can be programmed to complex tasks in a sequence of steps repeatedly without little or no human supervision.

With the advent of micro chips made of silicon the mammoth machines of the previous years occupying a large room has evolved into tiny sleek machines fitting a even a small suitcase. It is found everywhere, in our schools, workplaces, homes and even leisure places. The proliferation of computers into our lives leads to the question whether these computers giant and sleek adding machines or are they electronic brains? Will they ever be able to learn, speak, and even create their own kind? We know that they are not able to do so now but in future will they be able to do the above activities on their own? This paper looks into the details of cyberevolution and tries to answer the question of whether the computer will gain an upper hand on humans in terms of intelligence, common sense, etc.

Humans and Computers: A comparison
�Human beings are a group of species that took around 3million years to evolve. The human brain is a complex organ composed of around a 100 million neurons capable of doing a variety of calculations. The speed of the human brain is such that it can calculate around 2000 trillion instructions per second. Today�s computers are slow compared to the human brain. It is in this context that �Moore�s Law� on processing time comes into importance. It is stated that the computer�s processing speed doubles in every 1.5 years. Therefore if the processing speed of a computer today is such that it can calculate 40million instructions per second then by the beginning of the next millennium the speed would have reached 80 million instructions per second and by 2002 it will cross the billion mark. The processing speed of the computer increases exponentially. So it can be seen from the above that the processing speed of the computer will exceed that of human beings by 2030.� Speed at the beginning of 2002 will be 1billion instructions per second. So to reach 2000 trillion (1trillion =100billion) instructions per second it will take another 27 years from 2002 (=2030). (218 = 262144 there 18 cycles are needed to reach 200 trillion, but each cycle is 1.5 years. Therefore years needed to reach 200 trillion = 18 *1.5 = 27 years). But a higher instruction processing speed will not make the computer more intelligent than the humans.

�If we compare computers with humans the difference is so striking. The computers are made in a production line where each computer is made exactly similar to every other computer, but the humans procreate themselves. The computers cannot reproduce themselves. Today�s computers are able to amass large amount of information and store the data in it�s memory, but it is not able to link one data to another or learn from each data as the humans.

Present and Future
Human beings were able to develop artificial intelligence programs so that the computers could take decisions on their own and learn from the contingencies that are experienced during the execution of the tasks. Artificial intelligence researches has developed programs that could listen to spoken sentences and grasp their meaning, read news and other forms of text, and even talk with someone and write poems. 'Deep blue', an IBM Computer that was developed by IBM was able to beat the World Champion in Chess, 'Gary Kasporov'. Deep-blue had in it programs that would teach it how to behave and play in response to each move made by the World Champion. Now deep-blue is used for weather forecasting and other high performance area.

May be in a 100 years time there will be an intelligent electronic species� working as partners with human beings. �We will carry these small intelligent machines with us everywhere we go. We will just pick them up, tuck them under our arms and go off to attend business. They will be �Artoo-Deetos� without wheels, brilliant, but nice personalities, never sarcastic, always giving you a straight answer. They will be little electronic friends that can solve your problems�. These were the words of �Jastrow� in his book �The Thinking Computer� about the future of computers. Researches are going on in several universities and government institutions where great human brains are put to work to develop highly intelligent machines.

Intelligence as advocated by psychologists cannot be achieved in a single step. High level of intelligence is achieved only through learning and experience. Consider a newborn child, it doesn�t have any form of intelligence or common sense at birth. But as it grows its level of intelligence and common sense increases. Just like this no computer or computer program can be made to deal with all kinds of emergencies at the first time. The contingencies are difficult to model. Intelligent machines can be developed only through a period of time. In the process of developing these highly intelligent machines, human beings will lead the computers by hand, teach them until they reach their own intelligence level, after which they proceed to surpass humans in intelligence, problem solving and decision making capacity and even common sense. The question is when will this happen? In my view the speed in which the technology is advancing the day when the computers will surpass us human beings in intelligence and common sense will come in around 50 to 55 years.

But answering this question needs a look into attitude of human beings towards something more intelligent than he in his own environment does and the use of these intelligent machines. The use of intelligent computers will depend on the following questions.

1. Will it be economically feasible to substitute computers to do the task the human is doing?
2. Can computers be perfected to such an extent that it is able to reproduce exactly the human behavior?

From time immemorial humans had a tendency to create his own species by means other than natural procreation. This tendency has lead to the creation of legends like the �Pygmalion�, who created a statue of ivory to which �Aphrodite� gave life. Man has created idols that looked like him, which was immobile but was supposed to have life. It is in this context that technical progress is now taking place, technical progress that produces computers, which are more and more powerful every year, it is not surprising that human beings are tempted to attribute to them some sort of independent living existence. Man always wanted to build creatures like him and computer is no exception. In the present age man is trying to develop a machine that is as intelligent as him or may be more. We are heading towards that. The intelligent computers of tomorrow will not only be able to accumulate data and store it but it will be able to associate newer data to the data that is already stored in the memory. It will be able to think and learn. But it will always be under the human control and will be a faithful �electronic pet� of the humans.

Artificial Intelligence and Cyberevolution
�Human beings came into existence around three million years ago. At the beginning humans were nomads wandering in the land in search of food. The age of agriculture came around 50000 years ago, then came civilizations, industrial age and now the age of computers. Technological progress has accelerated lately. A lot of new inventions are being made in the field of computers, nuclear physics, human biology, etc. There has been acceleration in the field of technology in the near years. If we look at the curve of technological progress made by man during his existence on the face of earth we can see that this curve has been growing exponentially.

From the very beginning of artificial intelligence critics have pronounced that the essence of intelligence lay in the ability of the organism to learn from experience, dooming any efforts to create intelligent machines by programming them. Based on this thought artificial intelligence programs have been developed that is able to learn and then build a structure on what it was able to learn in the past. But is this intelligence?

Intelligence is the ability of an organism to learn, think and act according to conscience during contingencies. May be we could build a computer that would think and learn, but will it be able to act according to it's conscience in case of emergencies? Or does the computer have a conscience. To answer this question we should first know what a conscience is. Is it a single entity that can be modeled in a computer program? Is it something that is developed by learning? Or is just an instinct? Conscience is something vague that cannot be easily modeled in a computer program. But we can think of conscience as our whole system and the whole system can be broken down into parts. So we can model conscience in a computer program by thinking of conscience as made of different entities that communicate with each other and act together as a single system.

These programs will create large databases in which the action performed in each contingencies and actions to be taken in normal situations will be recorded. Whenever an event occurs the computer program will search the database for the action to be taken and takes action according to the database data. But what if a new event that has not happened in the past takes place? The same analogy of human evolution can be used here also. The computer will take some action by correlating this event with similar events that has taken place in the past and try to learn more from the result of the action. In this way the computer is able to learn more from the actions that it has taken and the result of those actions. These databases can be transferred from one computer to another, just as in the humans where knowledge is passed from one generation to another these databases that have been accumulated in one computer can be passed to the next generation of computers.

Cyberevolution can be stated as the change of computers from a dump state to an intelligent state through several years of evolution. At the very beginning the humans will help the computer to learn and meet contingencies until they reach the intelligence level of humans and then the computers will pass the humans in intelligence and will have an existence of their own. Each generation of computers will learn something and it will pass on that knowledge to the next generation and the new generation will add some knowledge to the database and pass on to the next generation. So each generation of computers will get knowledge from the previous generation and will be more intelligent than the previous generation. In this way computers will evolve and will pass the humans.

The important thing to note here is that cyberevolution is both hardware evolution and software evolution. In this the evolution of the software will be faster and will lead to intelligent machines. So in this context the software part of the computer can be thought of as the conscience of the computer that will help it to act in case of emergencies. Moreover the software from one computer can be transferred or copied on to another computer. So the transfer of intelligence and knowledge from one generation to another is facilitated. Development of this type of learning programs has been undertaken and hopefully in the next 50 years we will have a machine that is intelligence enough to take decisions on it's own.

Intelligent Machines and its implications on Human Factors
�Human factors is the science of study of those elements which influence the efficiency with which people can use equipment to accomplish the functions of that equipment [16]. The elements that influence the efficiency of equipment use are,

1. The physical characteristics of the equipment to which the human must respond. The examples of physical characteristics of the intelligent machine are the arrangement of displays and the control panels, if they need to be controlled.
2. The physical surrounding in which the equipment must be operated.
3. Characteristics of the task these intelligent machines and humans are doing.

The intelligent machines have an existence of it's own. The question is whether humans will have to interact with these intelligent machines as he does with machines that are not intelligent. The intelligent machines don't need constant supervision to operate.If we look at the definition of human factors, which states that human factors is the science that studies the interaction between man and machines. The machines are tools that are used by humans to produce something. These intelligent machines/computers however intelligent they are will always be a tool that man will use, though different from other tools that he use. Then these intelligent machines will also come under human factors study. The existence of intelligent machines will always reduce errors due to variability. But will it be able to reduce human errors? Since these machines are more intelligent than humans it can always reduce human error if not completely eliminate errors. These machines will continually monitor human actions and this will lead to reduction in human error.

May be in future the subject of human factors will develop into a science that encompasses the study of both man and these intelligent machines in relation to their environment. Human factors will be more inclined towards the study of these intelligent machines than the study of humans, as these machines will be performing all the work that man is doing today.

Conclusion
�Computers will not be the same some fifty years from now. It will overtake the human brain in processing capacity in 30 years time from now and will reach a stage at which it will be able to learn from it�s everyday experiences by around 2050 AD. But the most important thing about the evolution of computers from a �dump� adding machine into a highly intelligent machine is that �the computers don�t evolve but the humans who make and design them evolve�. It is this evolution that permits the construction of ever-improving and learning machines.

References
1. Warnier, J. D. (1990) Computers and Human Intelligence, Prentice Hall Inc., New Jersey, USA.
2. Flores, F., & Winograd, T. (1986) Understanding Computers and Cognition: A New Foundation for Design, Ablex Publishing Corporation, New Jersey, USA.
3. Geroge, F. (1989) Machine Takeover: The Growing Threat to Human Freedom in a Computer-Controlled Society, Pergamon Press, Oxford, England
4. http://www.charlesriver.com/books/bhchap6.html (Paul, S. G., & Cox, E. BEYOND HUMANITY: CYBEREVOLUTION and FUTURE MINDS )
5. http://www.charlesriver.com/books/bhchap3-1.html (Paul, S. G., & Cox, E. BEYOND HUMANITY: CYBEREVOLUTION and FUTURE MINDS )
6. http://www.charlesriver.com/books/bhchap3-2.html (Paul, S. G., & Cox, E. BEYOND HUMANITY: CYBEREVOLUTION and FUTURE MINDS )
7. http://www.charlesriver.com/books/bhchap1.html (Paul, S. G., & Cox, E. BEYOND HUMANITY: CYBEREVOLUTION and FUTURE MINDS )
8. http://www0.pbs.org/wgbh/nova/robots/moravec.html
9. http://telobs.com/TelecomObserver/Revolution/
10. http://www.geneticprogramming.com/Tutorial/tutorial.html
11. http://www.geneticprogramming.com/AI/fml.html
12. http://www.historyoftheuniverse.com/futucomp.html
13. http://www.historyoftheuniverse.com/evolutio.html
14. http://www.usc.edu/dept/NASA/tm/1996/dec/cyberevolution.html
15. http://www.geocities.com/Area51/Corridor/8102/cyberevolution.html
16. Meister, D. (1971) Human Factors: Theory and Practice, Wiley-Interscience, New York, USA.