Make your own free website on Tripod.com


Explore the Secrets of the Paranormal in:

THE PHYSICS OF THE PARANORMAL
by Kenneth W. Behrendt

Click here for a FREE preview!

Click Here to Enlarge Cover Image


Find Answers to the UFO Mystery in:

SECRETS OF UFO TECHNOLOGY
by Kenneth W. Behrendt

Click here for a FREE preview!

Click Here to Enlarge Cover Image


Find EVEN MORE Answers to the UFO Mystery in:

THE HOW AND WHY OF UFOS
by Kenneth W. Behrendt

Click here for a FREE preview!

Click Here to Enlarge Cover Image


Answers for the MOST serious of ufologists:

THE NEW SCIENCE OF THE UFO
by Kenneth W. Behrendt

Click here for a FREE preview!

Click Here to Enlarge Cover Image


Know Tomorrow's Science Today:

ESSAYS FROM THE EDGE OF SCIENCE
by Kenneth W. Behrendt

Click here for a FREE preview!

Click Here to Enlarge Cover Image




Artificial Intelligence and Consciousness

by Kenneth W. Behrendt



Like most "Baby Boomers", I was conditioned from an early age to believe that actual working robots would be invented during my lifetime. These creations were standard fare in the science ficton novels and movies of the '50's and '60's, yet it is now almost a half century later and we still do not have anything that closely resembles what was promised back then in the crystal ball of science fiction.

While we do not have the "Robbie the Robot" or last decade's "Mr. Data" that would make this age old dream of a fully interactional mechanical man a reality at present, it appears that the soon anticipated advances in computer technology and materials science may, indeed, make that dream a reality eventually. Even now, the budding science of "nano technology" suggests that it may SOON (as in a decade or two) be possible to construct computers which will have the awesome storage capacity of the human brain without any of its biological frailties. It will be possible to actually "grow" these artificial brains in various chemical solutions and program them in seconds with a lifetime of learning and a completely formed personality. Unlike a human brain, processes inside these artificial brains will take place nearly instantly.

Obviously, the next step after the development of these artificial brains will be to house these wonders of nano technology is some sort of high tech android body. These can even now be fashioned from various silicone or organic polymer elastomers that are virtually indistinquishable by touch or appearance from human skin. Such an android body would have a central hydraulic pump that would send pressurized liquid silicone oil to hundreds of small slave cylinders located throughout its body. These slave cylinders would function like the muscles in a human body and allow for the smooth movement of the android's limbs. Hundreds of pressure sensors embedded in such an android's "skin" would instantly tell it how much pressure it was exerting on objects in its surroundings.

To power this first true android, it would seem that the soon to be developed fuel cells would be ideal. Thus, this machine would, like a human, breathe in oxygen containing air which would then be channeled into a battery of fuel cells along with a stream of hydrogen gas. As these two gases were reacted inside of the fuel cell battery, the electricity generated would power all of the android's internal equipment. The only "waste" created in this process is pure water. When the android was resting and only using minimal power, this water could be exhaled as vapor. However, after strenuous exertion, a large amount of water might be created and this would, when convenient, be emptied from an internal storage tank via a process that would mimic urination.

The structural details of such an android's body are, surprisingly, the least important features of such a device. The critical part that will make it all possible will, in fact, be the complexity of its artificial brain.

Most of the devices available today which are referred to as robots are, in reality, nothing more than high tech puppets. They are basically remotely controlled devices which can only perform tasks in their environments so long as such tasks are carefully monitored and directed by a human controller. Thus, when one sees "robots" being used by the bomb squads to defuse a bomb or rovers moving across the surface of Mars to study its geological features, one should keep in mind that these devices are merely extensions of a HUMAN mind. A "true" android may initially be given a task by a human being to perform, but all of the many complex sub-tasks involved in accomplishing the larger task must be performed completely and INDEPENDENTLY by the android. This is a critical condition that such a device must be able to meet before we can truly consider it to be more than just a puppet.

This now brings us to the subject of "artificial intelligence" about which there has been much debate during the last half a century. There have been several definitions of this including the one given by Alan Turning, a mathematician, philosopher, and early pioneer in the science of cybernetics. He once suggested that a machine could be considered to be intelligent if a human conversing with it over a teletype machine could not tell whether or not he was conversing with another human being or a machine. This is a good definition and can be extended quite easily to the android described above. That is, if such an android could be dressed up and, after being placed in a social situation, not appear to humans by casual observation to be anything other than another human, then that android could be considered to have artificial intelligence. In other words, we could also say that, upon casual observation, there would be nothing about the android's appearance, speech, gestures, or other bodily movements which would give it away as being a machine.

Of course, we ALREADY have machines that have a very high degree of artificial intelligence. I am referring here to the familar personal computer that almost every household now possesses. Through its complex operating system (whose program can contain tens of MILLIONS of lines of computer programming code!) and additional programs that are temporarily downloaded into its "volatile" memory, such a device can execute complicated calculations or other data processing tasks in a fraction of a second. With the right programs or software, a computer can process images made up of millions of data points and detect patterns in them that would be imperceptable to the human eye. There are even programs now in existence that can "recognize" SPOKEN words, analyze the meaning of those words, and give spoken responses that are appropriate under certain limited conditions.

However, all that has been achieved in computer science to date is still only about a fraction of a percent of what will be needed to build a true android. Computers will have to become much smaller, much faster, and contain programs of staggering complexity. Only the field of nano technology would seem to offer any hope of developing these.

Let us now consider what the advanced artificial intelligence of a true android might involve. This should give the reader an accurate idea of the complexity of the program its brain will have to contain and be able to use.

First and foremost, the android's brain must be SIMULTANEOUSLY aware of the operational status of ALL of the many hundreds or even thousands of components that compose its body. Thus, it's brain must be able to sense such things as the temperatures of its internal components, the hydraulic pressures in EACH of the slave cylinder "muscles" which position and move its metal skeleton, the angular positions of all of its limbs, the electrical current output from its fuel cell battery, and the pressure of the gaseous hydrogen fuel stored inside of it's internal tank. The brain must also have a sense of balance so that, if the android starts to fall over, it can take immediate action to compensate for this and remain in whatever position it is supposed to retain. All of these various sensing and adjusting processes that will serve to maintain the android's internal homeostasis will, no doubt, require that a considerable portion of its artificial brain's progamming be devoted to them.

Next, we must consider the android's senses which, as in humans, allow this device to perceive various objects and energy sources in its immediate environment.

The sense of touch can be easily created by embedding hundreds of tiny piezoelectric pressure sensors in the plastic skin of the android along with tiny temperature sensing resistors. Each such sensing unit will provide a sensation of touch to an area of skin of about 1 square centimeter. ALL of these sensors will have to be connected by fine wiring to the machine's brain. As in a human being, their function will be to tell the brain if it is applying too much force as the android's limbs attempt to move objects in its environment. Temperature sensing is critical in helping the android avoid potentially dangerous environments and contact with fire.

The senses of sight and hearing are, perhaps, the easiest to achieve because we already have a technology that can produce high resolution digital cameras and high fidelity microphones. Of course, the data streams coming from these miniaturized devices will have to be rapidly analyzed by the android's nano technology brain so that it will be instantly aware of what objects are located in its immediate environment. The programming for the android's eyes will be further complicated by the need to provide the android with stereoscopic vision. Not only must all objects in the images from each camera eye be identified by their shape, color, texture, etc., but their parallax shift with respect to background objects must be measured so that the brain can determine how far away the objects are from the android's body.

The senses of smell and taste will be the most difficult to equip an android with. Although our android will not need to eat food or drink beverages, it may occasionally pretend to do so when in social situations with humans. This material would later be discharged, undigested, from a storage sack within its abdomen. It seems that, again, nano technology will be able to eventually provide sensors that can detect various aromas and flavors. These will be super miniaturized versions of the military "sniffers" used to detect the presence of toxic chemicals in the air during chemical warfare.

As was noted above, when this android is given a task to peform by a human being, the robot must be able to break that task down into a finite series of sub tasks and perform each of these successfully in a logical and efficient sequence in order to finally complete its main task. The android must be able to follow either verbal or WRITTEN instructions, be able to ask for additional information if needed, and be mobile enough to ovecome all barriers in its environment that might hinder the performance of its current task. Most importantly, the brain of a true android must be capable of LEARNING or acquiring new informaton about its environment. Although it will initially be programmed with a large amount of data concerning objects on earth, the physics and chemistry of everyday situations, and the various rules of conduct in human society, a true android's brain will still have to be able to ADD to this database when it comes across objects and situations with which it is unfamiliar. These abilities will, most likely, represent the ultimate level in artificial intelligence that can be achived by a machine.

The question quickly arises as to whether such an android would have consciousness as the term is generally understood by humans. The answer to this is rather complex. Obviously, the android's brain will be aware of objects in its environment and of the operational status of various parts of its internal mechanisms. However, in order to be conscious, it must also have "self-awareness": that is, the android must be aware of itself as SEPARATE from its environment. This ability will naturally to be part of its complex programming because all distances to surrounding objects will be computed with respect to the position of the android.

Consciousness for humans also requires the ability to engage in IMAGINATION. Thus, we can engage in day dreaming in which we "call up" various mental images related to problems we are trying to solve. We then alter these images in different ways as we try to predict how the real objects that the images represent might behave in various situations. Often this process leads to novel solutions to the many problems of human existence.

We see from this that humans, whether awake or while sleeping and dreaming, can produce a kind of virtual reality representation of the external world within the hidden neural circuitry of their brains. Any robot who claims it is conscious would have to be able to basically do the same thing. Its brain would have to be able to generate a virtual reality model of its external environment within its nano circuitry and then manipulate the object images of that model to see if any novel solutions related to the performance of the android's current task were produced. In other words, our android would not truly be conscious unless it had the ability to exercise CREATIVE imagination.

Some readers of this article may feel that a machine could never achieve the kind of consciousness that humans possess. The final reality of the situation may be that androids will, eventually, achieve a kind of "super consciousness" that will actually SURPASS that of humans! Again, all of this becomes possible IF the promises of nano technology can be kept.

An android would be able to nearly instantly identify objects in its environment by measuring their various sizes, shapes, and color patterns and then comparing them to a catalog of stored object parameters within the android's brain. If the object was not in this catalog, the android would be able to access a worldwide data base using wireless satellite technology. This same technology would also let any two androids or GROUP of androids communicate with each other via a special binary android language. Thus, a problem that might be beyond the analytical capabilities of single android could be submitted to a group for analysis and recommended course of action. Humans already do something similar to this when they read and respond to posts on internet forum sites.

In attempting to perform a task, an android could literally contemplate THOUSANDS of possible ways to complete its assigned task. It would then only try to act on the methods which were the least expensive or least energy consumptive. To an human observer, such behavior on the part of an android would appear to be highly creative. Indeed, it seems logical to expect that androids would eventually be inventing most of the devices used by humanity. Eventually, androids could even be assigned the task of improving their own kind by designing more advanced nano technology brains that could hold an even more complex personality program. This personality program itself could even eventually be created by android programmers working under the careful guidance of humans! Such a complex personality program might be customizable by the robot's owner so that the android would be able to get along better with its human owner. The owner of an android could take a personality test and the scores from its various sections would be used to suggest how to set the various personality parameters of the android brain's personality program so as to assure maximum rapport between human and machine.

There are predictions now that the first robots in the human home will be in the form of robot pets. Because a pet robot is smaller in size than a full sized android, the cost will be less for the owner. Even a small robot pet might be able to perform many useful tasks around one's home. It should even be possible to construct robot pets that would have near human intelligence! This will take some getting used to on the part of humans, but robot pets capable of human speech that could hold an intelligent conversation with their owners would be an asset to lonely people who would normally keep a live pet for companionship. For older people, such a robot pet, perhaps in the form of a dog or cat, could monitor the owner and call for help in the event that the elderly person was sick or injured.

When the day finally arrives on which we have independently functioning androids walking the streets of our towns and cities as they go about performing their assigned tasks, some humans may be bothered by status of these machines as servants in our society. However, one must, I think, always keep in mind that an android will only have the feelings or emotions that its human owner decides to allow it to have. Thus, it will not be angry that it must work 24 hours per day, 7 days per week. It will not feel that it's rights are being violated or that it is being taken advantage of. It could only have such feeling IF its human owner decides for some purpose to allow it to have those feelings. One could just as easily program such a machine to express happiness and gratitude for being assigned to continuous labor of the most menial sort! Such happiness would be absolutely genuine as far as the android was concerned.

The first truly independently functioning androids will, unfortunately, be VERY expensive. This cost will only be justifiable if they are required to perform tasks that are too difficult or dangerous for humans to perform. Hopefully, with mass production techniques, it will be possible to eventually lower the cost of an android to the point where the average person can own one. Such a device could provide companionship or perform various domestic services around the home which could include such things as cooking, laundry, home repairs, lawn maintainance, etc. In short, all of the chores that humans tend to avoid could be assigned to an android.

Although humans will acknowledge that androids are intelligent and even conscious soon after their appearance in our lives, I do not believe that these machines will ever be LEGALLY considered as another lifeform. They will therefore, ultimately, not be accorded any more rights than one's automobile or home computer. This status will not be offensive to the androids because, as was stated above, they can be programmed to willingly accept any role assigned to them in human society. In general, this role will be to serve the needs of their owners.

Of course, as with any potentially dangerous piece of equipment, it will be mandatory that strict safeguards be incorporated into them which will prevent them from being misused. Thus, no robot or android would be allowed to follow any instruction or continue with any task that its artificial brain's analysis indicated might lead to the injury or death of a human being or other higher lifeform. There would be strict penalties for any human being who attempted to bypass this safeguard.

However, one wonders how such an android would react to a physical attack upon its "person" by a human being. An android would be an expensive piece of equipment and it would seem logical to give it some ability to prevent itself from being damaged or destroyed or stolen. But, what if the OWNER of the android had decided for some reason to destroy the machine? Perhaps in these cases, the android would take immediate evasive action to protect itself if attacked by any human other than its owner, but would submit to an attack by its owner.

What might an android do if its assigned task was to protect its owner from some other human being? In such cases, the android might only be allowed to use non-lethal force or weapons to perform its task. In these cases, the android would not be allowed to use firemarms, knives, or clubs, but would be restricted to the use of such things as chemical, optical, or electrical weapons.

The behavior of androids in such situations will pose a variety of moral and ethical problems that will have to be ironed out by the legal system and the companies that create the programs for these androids. Completely satisfactory solutions will not come over night, but may take many decades to achieve.

While one can easily imagine androids performing many of the menial tasks in a society that tend to be boring, dirty, or dangerous, their amazing intellectual abilities would also suit them well in performing many of the "routine" tasks that are now the done by various professionals.

Thus, I can envision hospitals staffed by android doctors and nurses. A surgical team of androids would be able to perform the most delicate of surgeries in a matter of minutes and, thereby, greatly reduce the risk to the patient from the comatose state induced by anesthesia.

Android medical researchers might be put to work to develop some new lifesaving drug or procedure. Unlike humans, the androids would be able to work on the problem around the clock, day after day without the need to sleep or eat. They would only occasionally stop for less than a minute to replenish their internal supplies of hydrogen fuel. Once the company that owns the researcher androids had paid for them, that company would have a source of highly skilled labor that would virtually work for free. Of course there would be the very occasional cost of maintainance for each android. But, not having to pay human researchers high wages, would help greatly lower the cost of new medications. This drastic reduction in the cost of development would help make these drugs widely available to even the poorest of countries.

It is easy to extend this scenario and predict that androids will eventually make human labor unnecessary. However, because of the high cost per android, I see little chance of this happening anytime soon. People will always want to make sure that somewhere in any organizational hierarchy that directs and controls android workers, there will be human beings in charge. Thus, no android or group of androids will ever be operating completely independently of human oversight. This will prevent the possibility of some situation coming into existence that might pose a threat to humanity. If such an unforeseen situation did ever occur, then it is important that human beings be able to override it and shut the androids down until the problem could be extensively analyzed and corrected.

So, when can we expect the arrival of the kind of android envisioned above?

Predictions in the fields of science are risky things to make, because, more often than not, they tend to be wrong or way ahead of schedule. I remember that during the '60's the futurists were predicting that we would have lunar cities and regular travel to Mars by the end of the 20th century. Now that we are at the beginning of the 21st century, these things are now being predicted for the middle of our century. I am currently of the belief that, if we are restricted to rocket technology, even these latest forecasts are extremely optimistic and ahead of schedule.

When it comes to android technology, however, I would be very surprised if the machines described above were not perfected by the middle of this century. The structural materials needed are already available, a miniaturized, high efficiency hydrogen fuel cell power plant is almost here, but, unfortunately, the nano technology needed for the android's brain is still only in its embryonic stage of development. This last element is an absolutely critical one that MUST be developed in order to make lifelike androids an everyday reality. Present day microprocessors simply do not have the data processing power needed to handle the enormous amounts of data that would be entering the artificial brain of an intelligent and conscious android.

In the future, with the help of android workers, humans will be able to construct a world that is safer, cleaner, and more comfortable and fulfilling for all of Earth's people than is our present one.

(Note: this article completed August 15th, 2004)








HOME
COSMIC VAULT GIFT SHOP