visitors since
1999 09 10

© 1998 - 2008 Keith Wiley

All material on this website is
copyrighted and may not be
used without first obtaining
permission from the author.

Thank you.

Disclaimer

Many of my essays are quite old. They were, in effect, written by a person who longer exists in that my views, beliefs, and overall philosophy have grown and evolved over the years. Consequently, if I were to write on the same topics again, the resulting essays might differ significantly from their current versions. Rather than edit my essays to remain contemporary with my views, I have chosen to preserve them as a record of my past inclinations and writing style. Thank you for understanding.

April 2001

Robots and Mind Uploading

The near future will knock your socks off, the far future will be incomprehensible

Brief Description

Presently we are seeing the first stages of a robotically oriented society. In ten years this preliminary stage will be indoctrinated into our culture as robots start appearing in all corners of life. In twenty years simple robots capable of performing work (as opposed to today's dinky toys) will be ubiquitous in the home, the office, and industry. All jobs will be in the middle of the transition to robotic automation. In thirty years our society will have undergone a major shift in the workforce as robots take over most jobs. In forty years, robots will be capable of performing any human function. In fifty years they will be immeasurably superior to humans at every conceivable function. The predictions just made may be off by up to ten years in either direction, but probably not more than that. Then the party really starts. In one hundred years, humans in their present form will be extinct while an entirely new form of humanity will have emerged. Yes, just one hundred years, possibly several decades earlier, maybe a few decades later, but this prediction is not off by more than a few decades.

Full Description

Sections:

Introduction
Computer minds: AI
Computer bodies: robots
Mind Uploading
Timeline
How to directly benefit from these predictions
Related links

Introduction

For the last several thousand years human technology has been increasing in complexity. However, it was only in the last century that we began to make machines that hinted of mechanized thought. People have been predicting that artificial intelligence was just around the corner for forty years and so far they have always been wrong. I believe this will change very soon. While we are busy making computers more intelligent, two other disciplines will be occurring alongside the computer revolution. One, robotic bodies for intelligent computers to inhabit will also become increasingly adept. Two, neuroscience will begin a merger between computers and brains.

Computer minds: AI

As far as raw computer intelligence is concerned, there are some very good reasons to believe the tide is about to turn. For one thing, in the last ten years or so, a new approach to artificial intelligence has become popular that wasn't embraced in the past. This approach consists of replicating biological methods for developing advanced, complicated computer programs. This field is called artificial life and encompasses many different subcategories including self-learning neural networks, genetic algorithms, subsumption architecture, emergent complexity, and several other disciplines. By allowing computers to develop intelligence, consciousness, and individual behavior on their own instead of having to be programmed with these attributes outright, I believe we will finally succeed in attaining the goal of artificial intelligence. It has proven to be incredibly difficult to engineer a system as adaptive, robust, and complex as a brain, but the approach of evolving such a system is much more likely to succeed.

Another reason to believe computers will soon become very intelligent is that within the next few decades the hardware that is required for human level intelligence will become available. It is pointless to predict a smart computer program if no computer available is powerful enough to run the program, but computers are approaching human levels of ability in terms of raw computational power. Once the hardware is available, we will most likely find a way to allow that hardware to express itself as a form of cognitive life.

Computer bodies: robots

What good is artificial intelligence if it sits in a box on your desk? Well, actually, it could be argued that a truly intelligent computer could perform many useful functions just existing as a box. However, it is physical robots that I personally find the most tantalizing. Presently, robotics is far behind computer science. There are two primary problems with robotics in my opinion. The first and most serious is that we have yet to develop an efficient power source. Most robots have a lot of extremely heavy batteries, and those batteries don't run for very long despite their size and weight. There are so many looming alternatives that I wouldn't know where to begin. I will list a few: fuel cells, solar, air (wind), water (flowing), fusion, chemical (like biological organisms use), geothermal, kinetic, flywheels, microwaves, and many more that I can't even think of right now, not forgetting good old fashioned batteries if we can greatly improve on them. Biology is fabulously energy efficient. Most animals can eat a small meal that comprises only a fraction of their own weight and can then function for days without food. There is of course great variety in the animal kingdom. Some animals spend practically all their time eating (pandas) and some go weeks without eating (snakes). The average robot is good for about ten minutes. Robots will have to find a way to store energy as efficiently as biology if they will ever have a real chance of competing.

The second major problem (and this ties in directly with the energy source problem) is that we don't have energy efficient actuators, methods of producing physical movement. The most common method of movement is the motor, an electromagnetic contraption that always produces a rotational motion. There are many other actuators including shape-changing alloys (metal wires which alter their length as their temperature changes), pneumatic (air pressure), hydrodynamic (liquid pressure, usually water or oil), and a couple others. These methods vary greatly in energy efficiency, but none of them are really very good compared to the muscles that biology uses. Until we find a way to produce motion using energy efficient methods (which takes into account the volume, mass, and energy requirements of the system as a whole), robotics will never take off.

Mind Uploading

Aside from artificial intelligence and robotics, there is a third field that will start showing incredible promise in coming decades. We will begin to integrate computer systems into our brains. We have already begun. In fact, this field began hundreds of years ago with the first hearing aids (in the form of small funnels) and the first eyeglasses. These were the first tools designed to augment our brains as opposed to some other part of our bodies. Such devices have become considerably more invasive in recent decades including hearing aids inset into the outer ear, hearing aids hidden behind the eardrum inside the middle ear, contact lenses, laser eye surgery, and the rudimentary beginnings of spinal injury repair. In addition, neuroscience benefits from some of the most amazing scanning devices in all of science. Our ability to learn about the brain through electrodes, PET, CAT, and MRI scans is increasing in both spatial and temporal resolution on a constant basis. We will continue to use computerized devices to fix neurological problems using increasingly invasive procedures.

Eventually we will begin to perform such procedures, not for the purpose of fixing defects, but for the purpose of enhancing already healthy abilities. Computers will become more and more integrated into our brains until we will eventually dispense with the old-fashioned goo that is our biological brains and move forward at accelerating rates with our fully artificial brains. We will have, in essence, become the robots. This concept is called Mind Uploading, and there are several approaches. A more fantastical and more generally recognized method than the one I just described consists of simply taking an ordinary brain, scanning it by either destructive or nondestructive means, and then producing a simulated copy in a computer, entirely free of the original biological body. I find this scenario to be less likely in that I believe the necessary technology would arrive long after we have been augmenting our original brains with computers, but in theory it is just as explanatory an example. Not only will we be through with our brains, we will be equally bored of our simplistic and overly fragile bodies. Robotic bodies will be our minds' vehicles of choice, no doubt.

The following timeline is an approximation of what I believe the course of events will be over the next few decades. I allow myself a latitude of approximately twenty percent. In other words, if a predicted date is ten years in the future I permit myself an error of plus or minus two years. Twenty years in the future permits me plus or minus four years, and so on. The reason for a fairly loose time-constraint is that unpredictable political and social factors may sway the emergence of particular technologies by several years. For example, in the United States, the difference between the election of a technology-embracing president and a technology-denouncing president could shift the entire schedule by as much as four years.

2000	It is presently mid-2001 and I use the year 2000 as a round number to illustrate the present state of AI, robotics, and Mind Uploading. A desktop computer has barely an inkling of human-level intelligence, perhaps one millionth of our brain's ability. This is insufficient power to perform complex generalized computations. However, this is sufficient power to perform complex specialized computations. In other words, we can't make a general-purpose intelligent computer, but we can make a computer or robotic system that can recognize human faces, dictate smoothly spoken speech, steer cars down highways unaided at highway speeds, operate partly unsupervised on Mars, play humanly unbeatable chess, run approximate weather prediction models, weld cars together, perform astonishing feats of optimization that are untouchable by human abilities, act like fairly amusing although easily recognizable artificial pets, mow your lawn, deliver meals in hospitals, and myriad other examples. Many of these systems are prototypes and are not ubiquitous in industry or otherwise in our daily lives but they haunt of the future that is to come.
2010	While some robots in research labs will have been using legs for over twenty years back to 1990, most robots will still be using wheels, which are far more energy efficient than legs. Wheels can't traverse stairs or rough terrain very easily, and robots will be limited in this respect. House-vacuuming robots are a good example of what the average consumer may have available at this time. They will be fairly simple, yet certainly useful in a particular domain, and more to the point, they will be affordable. Cars may have failsafe mechanisms that take over the wheel when the user doses off, but they won't be driving themselves yet. We will probably have language translators that take in and produce smooth-sounding speech. Robots will be showing up in places that most of us won't see them, such as warehouses and factories. The job-crunch will have begun as some jobs become doable by robots for a cost that is competitive with human wages. Robots can work almost around the clock, don't complain, don't strike, don't need vacation or sick-leave (although they will break down). Consider this: a robot costing a company $20,000 up front and $3000 yearly maintenance thereafter will outsell all but the lowest paid human workers if you think about it.
2020	Robots will have a more general, although still somewhat simple, applicability. Some robots may benefit from the use of legs if their function requires it but will probably still have wheels also (because wheels are still more energy efficient provided a flat surface). A general purpose house-robot can vacuum, straighten up a room, put away (possibly wash) the dishes, find and wash the laundry, mow the lawn, change lightbulbs, sweep and wash the floors, wash windows, etc. Package delivery robotics may be in full swing at this time. Small packages will be delivered throughout urban areas by an army of small robots instead of by a small truck-team of humans as was done in the past. As a result, package delivery will speed up enormously, while probably becoming much cheaper at the same time. Cars may be fully self-driving, but I believe that the safety concerns may push this specific application further back than should be necessary. Perhaps not though. Practically all surgical medicine is performed by robots, which in most cases aren't even being directed by humans. Robots will be stealing more jobs, especially basic service industries like fast-food (in which present day employees must perform basically like robots anyway), janitorial jobs, and most cashiering jobs. It is unfortunate that the first people to lose their jobs will be lower-income people, but I believe this is an accurate prediction. Some form of computerized brain augmentation will be occurring. Perhaps infrared/ultraviolet cameras will feed their vision information directly into our retinal system. We might enjoy similar augmentations for hearing new frequencies of sound. It may be possible to record the neural pattern of a given experience and then play it back through neural stimulation. This won't allow direct virtual reality yet, but may allow sensations and feelings to be induced.
2030	All forms of transportation are "hands-off". General purpose robots will be even more adept. You can send your personal robot to the store to fetch the groceries, or it can work on car and home maintenance for you. More and more jobs will be going to the robots: public transportation obviously, more service industries like the basic utilities of plumbing, electrical, and gas. Factories will be entirely empty (of humans). Banks, department stores, most large retailers for that matter, government offices, and other places where the customer is basically in-and-out without much human interaction will have abandoned human employees as they cost too much and make too many mistakes. Places with a human touch, like libraries, schools, mom-and-pop retailers, small motels, restaurants ,and such will still have a human somewhere in sight when you walk in the door, not because it's necessary for the job but because people crave interaction with other people and robots will not provide ample replacement in this regard yet. It is an interesting question to ponder where all the humans will have gone. I seem to be describing a desolate wasteland of sorts. Basically, I believe this demonstrates the recreational lifestyle that we will be adopting, a lifestyle in which most of our time is spent on pleasurable pursuits instead of financial ones. Economics will still exist and money will still be important, but we won't be working in the same kinds of jobs as in the past (and certainly not for long hours). What kinds of jobs will we be doing? As far as I can tell, the only domain left is robotics research. Just kidding. We will be more artistic and less industrial, more creative and less menial, more mental and less physical. I will leave this issue as a vague suggestion however.
2040	Robots will be performing every single job at this point, including upper management. Companies may be "owned" by humans, but computers will run everything. We will simply collect stock dividends to support our hedonistic desires. A major problem faces humanity now. Robots are intelligent and conscious enough that an ethical dilemma arises about the continued forced labor of such machines. Meanwhile, our brains will be undergoing substantial augmentations. We will enjoy plug-in math processors that can perform calculations for us literally "at-will". We will likely have something akin to perfect virtual reality in which a virtual world is projected directly into our neural system and our muscular movements are redirected into that virtual world. It is an interesting question whether we will find any interest in the physical world when this becomes possible.
2050	Whether really versatile artificial intelligence or fully uploaded human brains occurs first is a huge question. The answer to this question may alter the future of humanity in practically orthogonal ways. If AI comes first and has the power to enforce its rights, and if we don't tread very lightly on such matters and treat such demands with respect, we may be eradicated. If Mind Uploading comes first, we may perpetually enslave conscious, feeling machines for all eternity into doing our menial work for us. The truth will probably lie somewhere in between. I believe AI is easier than Mind Uploading, and resultantly I believe the machines will arrive before we become the machines. It is difficult to predict with any accuracy whatsoever what the world will be like at this point, and this is only fifty years in the future. Technology will have reached the farthest corners of the Earth. It will be impossible for any society, even the most conservative ones, to avoid the technological advancement of humanity. It should be entirely possible for world hunger and disease to be vanquished even in the most remote areas. Universal language translators will have turned our world into a more unified, singular population. The production of electrical energy will be cheap and will require small physical devices instead of huge power plants. Consequently, electrical power, the computing power that comes with electrical power, and the robotic power that comes with computing power will all exist on the Serengeti, in the outback, in the amazon, in the arctic, etc. Basically, everywhere there is a human, there will be advanced technology to aid that human. Of course, what it means to even be human is debatable since robots will possess superior intelligence and we will have a variety of devices implanted into our bodies and brains.
2100	Assuming we didn't bring destruction on ourselves through oppression and bigotry of robots, we will be enjoying an existence we can't even imagine today. Our minds will run on entirely mechanical "brains". They will run at far faster speeds than they presently do. Our memories will be infallible. We will have tremendous calculation abilities. We will have unimaginable modalities in vision, hearing, smell, taste, and touch, in addition to senses we don't even have presently, such as the magnetic sense that birds appear to possess. We will be wonderfully artistic creative beings with the universe at our disposal to explore and experience. Every ounce of dead matter will be converted into living computing machinery as we will literally give life and consciousness to the inanimate universe. When an individual wishes to, he or she will be able to share in a group consciousness that will embody all "humanity". Any experience will be sharable, including experiences of emotional and sensory intensity (let's call it sex), despite an ironic lack of need for sexual reproduction anymore. Anyone wishing to escape civilization will be able to jet off in a random direction into uncharted space. There will be literally, practically nothing that we won't be capable of doing. There is a little further explanation of what sort of robotic form we may take in my other Mind Ramblings topic, Religion's Impending Death, in the section where I discuss the future and immortality. There is also a hypothetical scenario involving virtual reality mentioned in my other Mind Ramblings topic, Aliens, UFOs, SETI.

How to directly benefit from these predictions

One last point. I believe that it is possible for people in the young adult or younger years today to survive long enough to experience even the extreme time-periods in the above predictions through coming advances in medicine, longevity, surgery, organ-replacement and life-extension. In other words, I hope to be alive in 2100, fully robotic, virtually immortal, and practically a god. If ever you needed a reason to start thinking about your health, this is it. In the past, regardless of your health, you knew you were pretty much walking the same path as everyone around and before you. A few decades one way or the other wasn't going to affect anything. Now things have completely changed. If you die a decade before Mind Uploading gives us the ability to transfer our mind between physical substrates with ease, you won't have lost a mere few years of life. You will have lost countless millennia of expressive experience that the rest of humanity will go on to dream into reality without you. The potential gain is unfathomable and the potential loss is equally unbearable.