Technology and Evolution

Essay by: Muhammad Hozien

 

G. B. Dyson, Darwin among the Machines: The Evolution of Global Intelligence (Reading, MA: Perseus Books, 1997). ISBN: 0-7382-0030-1.

Here is a list of Book Reviews:

FoRK Archive: Book Review By: Lisa Dusseault

BookPage Review by: Michael Pellecchia

Book Review by: Denis Susac

EVOLUTION: n. 1. A gradual process in which something changes into a different and usually more complex or better form. 2. a. The process of developing. b. Gradual development. 3. Biology a. The theory that groups of organisms change with passage of time, mainly as a result of natural selection, so that descendants differ morphologically and physiologically from their ancestors. b. The historical development of a related group of organisms; phylogeny. 4. A movement that is part of a set of ordered movements.

On Intelligence and the life of machines…

The usage of certain words colors how we view technology. By describing the artificial, i.e. mechanical, in biological, or living terms, we get a view that technology has a biological life of its own. We in turn view ourselves as creators on the par with the Creator.

Meanwhile in describing the biological, human in this case or any living matter, in functional mechanical terms, we view life in a mechanical, non-living way. Humans become nothing more than a machine. The creator is reduced to nothing more than a ‘great’ Architect (p.20), a mere mortal. We blur the line between the creator and the created.

Such role reversal [where the created becomes a creator] is quite arrogant on our part. Human inventions can be traced back to what is found in nature. Carl Sagan said in ‘Cosmos’ that to bake an apple pie from scratch we have to create the entire universe. We are merely cooks combining ingredients to create a better pie than our neighbors.

While the use of such metaphors helps us in understanding i.e. the function of biological process and/or mechanical process- that imitates a biological process- making any more inference from such a metaphor is overtaxing it.

Wrong Metaphor:

Over taxing a metaphor will lead to wrong conclusions. This way we are creating a reality that is of our making. This is like building a mechanical arm that imitates a human arm. Amazed with our product we then describe the functions of this arm in biological terms. Then in a leap of faith we describe the long arm of the law ‘tyranny’ as a mechanical arm.

One can not infer from such a thing that the long arm of the law is imitating the mechanical arm that crushes everything in sight. Blind tyranny becomes equal, in our imagination, the mechanical arm. Mechanical arms ‘iron fist’ are then evil and tyrannical. We then take this inference a step further and say that there is no distinction, or only a subtle human contrived distinction, between artificial and natural.

Such a use of metaphor will help us understand the either biology or technology on a certain level of complexity. What I am saying is that it is only the truth at one level. Life looks mechanical on the surface, if you look at it in a shallow and myopic view. It maybe a tool that we could use to help the novice understand life or technology on a level that they could understand.

Our language tends to be technical and terminology oriented, however when we use the metaphors to explain things to the lay person we should not get carried away by our metaphors.

Is an intelligent Machine alive?

If we say that other intelligence does not have to be like our intelligence, i.e. requiring life. What are we really saying? Are we saying that mere calculations constitute intelligence? Does the computer, which is basically a machine that does many binary calculations, have intelligence? Does the abacus have intelligence, does the calculator? No!

The ‘net’, which is a collection of wires that route ‘digital signal’ (electromagnetic current) connected to digital storage processors is not alive. What we interpret this digital signal is irrelevant to the net. It could be voice, text, video or gibberish is of no consequence to the net. The more signal going through the wires ‘the net’ the slower it gets. Whatever signal goes through if it is formatted correctly (following the standard IP protocol) goes thorough to a specified destination before getting on the net.

Our language makes us think of advanced machines as intelligent and even living. We define a system that has feedback as ‘intelligent’. Further when it has ‘intelligence’ we describe it as ‘living’. These are nothing more than metaphors that we use them to help us understand the world in which we live in, to humanize the machines, and for an occasional dramatic effect.

Killing the Machine:

When a system is made we say it will go live on such and such a date. All we are saying is that the ‘start date’ to start a process is such and such period. We use live to give it a sense of urgency and perhaps a little drama.

Further our usage of the term living does not denote the full meaning of life. When we ‘upgrade’ a system or ‘retire’ an old one. By doing so we do not mean that we are killing a computer. We do not go to jail for committing homicide against old IBM PC XT. We do go to jail if we do disrupt the network of e-mail systems. Even the original computer, colossus or ENIVAC, that was built no one was sorry when it was shut down. We have better and faster computers and that is all that matters. New tools replaced the old.

If someone offers you that they will give you a computer that is 1000 times faster that your computer and will help in doing your mundane chores, you would take the offer no questions asked. If the government, or some other agency, said we are going to replace the entire ‘net’ with a much faster and improved optical ‘net’ no one would feel sorry for the old copper wire ‘net.’ There is nothing sacred about the tools that we have to achieve our ends.

Working with the tool:

We do adjust the way we live to accommodate a tool. This accommodation is done with regards to what convenience it will offer us in return. I will learn how to type on a PC if allows me to do more things than I can do with a pencil. I will carry a computer home if it will save me some time at work. I will be giving up comfort for the gains that I will reap.

We do not feel that we are under any moral obligation to keep the old machines running, the same way that we feel that we are morally obligated to care for the elderly, animals or the environment. Perhaps the only moral obligation that we feel when disposing these machines is where to dispose of them.

The sale and success -(having sold more than 3 million devices) of the Palm pilot PDA (Personal Digital Assistant) a device that is much less powerful than a Personal Computer- is an example of how we sacrificed the functionality of a PC for the sake of comfort and portability. This device has taken the most used applications and built a device around it.

Losing the big picture:

It is true that sometimes we lose site of what we are trying to accomplish. We get bogged down in the details and get lost in the tree to lose sight of the forest. This happens with any complicated tool that we use that requires tremendous amount concentration, it does not have to be high technology.

This is also true of large construction projects that require the entire resources of a nation. It is not unlikely that some Egyptian noble stood back and took a look at the pyramids and said ‘we must be slaves to these pyramids and we are their servants.’ The pyramids provided a value to their society, whether it be scared burial sites or prestige, it was their tool.

On the development of Machine Intelligence...

Is a complicated machine, in general sense, a living organism? Do we see intelligence in an organized and highly complicated machine? Do complicated machines tell us something about its creator, inventor, and or maker?

Do we see a reflection of our own intelligence in the machinery that humanity makes. Are we projecting our intelligence on these machines or are the machines reflecting our intelligence that was used in creating them? If it is so complicated that we do not understand it do we readily recognize intelligence? (i.e. do we lose sight of it!)

George Dyson gives some convincing evidence and presents a good case for machine intelligence. What follows is a summary of his arguments.

Starting with Hobbes who saw society as self-organizing organism that possess life and intelligence on its own. Hobbes explained in his book ‘Leviathan’ that the state functions very much like an organism. There are many organs that the state has with money circulating from one individual to the other. The individual benefiting from money is like a cell benefiting from the flow of blood in the body. He saw the functions of society as a whole, which he considered to be one body, having intelligence –namely a mind of its own.

Samuel Butler believed in species-level intelligence. Butler believed that humans were reproductive organs of machines. Much in the same way as bees is part of the reproductive cycle of plants. Butler also saw that the telecommunications network as the central nervous system (Neural Network) of intelligent machines.

Butler builds on Hobbes’ idea of society having intelligence and foretold of a day when machines could have intelligence. He saw humanity as advancing the evolution of machines. It was not hard to see humanity as serving the machines. Since they required constant attention, it was like raising a baby. Those that work with machines at times refer to them lovingly in human and living terms. Stephen King wrote about a car that was served by a human in his story, Christine.

Dyson showed how machine intelligence advanced from the time of Hobbes to our day. Leibnez was interested in how the mind of the machine would work. His research would lead to the invention of the Turing machine, the forerunner of modern computers. Turing’s major contribution was the coding principles, which was the basis for the software concept, namely the step by step operation.

With the arrival of software the nature of machine intelligence would change from theoretical to practical. It was no longer machines in humanoid forms that would have intelligence. Computers that were large and bulky and required much attention displayed intelligence. It is their calculation ability that was impressive. With each successive generation intelligence of machines would be on the rise. It would surpass the power of any human to calculate at their speed.

Turing saw the relationship between machines and intelligence. He recommended that we start out with a program that would simulate the operation of the mind of a child as a first step. This would later evolve into the mind of an adult.

Von Neumann also stressed the importance of an evolutionary approach to machine intelligence. He believed that a computer could be built that would be as complex the human brain or neural network. This would be accomplished by evolutionary growth of self-reproducing artificial neurons.

Von Neumann for a time would lead the computer revolution and would support AI research at Lab in Princeton. He also had another idea, in which he was trying to predict the economic behavior, which lead to the development of fuzzy logic. Here computers would be allowed to make mistakes and adjust to them. This would lead to self-learning computer programs.

The idea of self-adjusting would be an important step in the development of AI. If a computer could teach it self and serve itself, it would be well on its way toward total independence from humans. Machines would no longer need humans to self-reproduce. It would not be long before the day in which humans will not be able to self reproduce without depending on machines.

The intelligence of machines does not need to be comparable to human intelligence. Human intelligence does not use logic for thinking all the time. We know from experience that our emotions, not to mention our physical condition and other external factors effect our thinking.

Perhaps even fuzzy logic is not a sufficient model for human thinking. The human brain contains an average of 1012 neurons that are interconnected by 1015 synapses in an area that is smaller than the size of a football. We are talking about a very, very complicated organ. The functions of the human brain are not still fully understood, much less its interactions among other organs. We have a long way to go if we are to try to duplicate human intelligence.

 

Essay by: Muhammad Hozien

 

Previous PageHome PageE-mail: Muhammad HozienNext Page

Hit Counter