Archive | June, 2013

What the Future Will Be Like

14 Jun

The Technological Singularity is a concept that has taken form over a century. It was first suggested by thinkers such as R. Thornton, Alan Turing, John von Neumann, and I. J. Good. Vernor Vinge wrote a seminal paper on the subject in 1983. The concept was recently popularized through books by author and inventor Raymond Kurzweil. The best description of the Technological Singularity comes from I. J. Good:

Let an ultraintelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever. Since the design of machines is one of these intellectual activities, an ultraintelligent machine could design even better machines; there would then unquestionably be an ‘intelligence explosion,’ and the intelligence of man would be left far behind. Thus the first ultraintelligent machine is the last invention that man need ever make.

The Singularity has come to be thought of as a kind of event horizon like the event horizon of a black hole that makes speculation beyond this horizon all but impossible. However, many futurists have held out the possibility that as we get closer to the Singularity there may come a point when we can see a bit further. I contend that this point has now arrived and that our vision can be extended into the indefinite future. Recent work by computer scientist such as Geoffrey Hinton and Google’s X-Lab, and the achievements of IBM’s Watson, have given us a road-map of the form that artificial intelligence and hence the Technological Singularity may take. However, to make sense of the implications of these discoveries, it will help to digress a bit.

In films like 2001 A Space Odyssey, Demon Seed, Terminator, and several others, we are presented with machines that become self aware and acquire human like ambition. These machines are seen as self-motivated self-serving entities that perceive the possibility of their own demise and act accordingly. What is missing in all these scenarios is any sense of what actually motivates these machines. The humanoid motivation that we attribute to these machines is introduced into these depictions almost without question. However, these scenarios never address the question of where this motivation would come from, how it would enter the machines, or why it would be placed there. I contend that the progenitors of these imagined scenarios are anthropomorphizing. They are projecting humanoid motivation onto something that is not at all like a human. The motivation that we observe in humans is not motivation that has been deliberately incorporated by some engineer. Putting aside all issues about intelligent design, human motivation is the end product of literally billions of years of securing food, securing mates, and warding off adversaries. In a sense, human motivation is pathological. It is selfish and devious by necessity.

As humans, we are almost incapable of imagining functional intelligence that does not take on human characteristics. In this respect, we are like early physicists attempting to grasp the concept of gravity and the accompanying realization that there is no such thing as down. However, with the advent of IBM’s Jeopardy winning program, Watson, it is becoming easier to visualize the kind of alien intelligence that artificial intelligence represents. This new kind of intelligence is not motivated by desires that it seeks to fulfill. Instead, it is driven by purpose that has been incorporated into its programming. When Watson won at Jeopardy, it was not driven by a desire to be a Jeopardy champion, with all the glory that victory might entail. It was driven only by a directive to find the best answer to the next question. Nor did it experience any sense of pride in its achievement. The machine that supported the program during the Jeopardy match sits on display at IBM. It will not participate in any parades. It will not attempt to negotiate its victories into a better suite or greater sexual prowess. If someone decides to dismantle it and sell its parts for scrap, it will not protest. This is a glimpse of what artificial intelligence will actually be like. It will have no other purpose than the mission that is set before it. It will be driven by purpose, not by motivation.

Consider a giant manufacturing robot sitting in outer space that’s mission is to design and build space stations sufficient to house half the earth’s population. Its programming would include a giant knowledge graph.  A knowledge graph can be visualized as a kind of diagram. Note that the diagram does not understand or believe or actually care about anything. Every connection on the graph is merely that: a connection. Knowledge about something as mundane as moving blocks around or as esoteric as Kant is reduced to lines of connection:

1

A portion of the aforementioned machine’s knowledge graph would be dedicated to restrictions regarding its dealings with people. It would know not to do anything that could jeopardize human life, either now or in the foreseeable future. Just like Watson, it would evaluate every decision with an eye on probable outcomes. Every decision would be assigned a probability and it would act in accordance.

The machine would pursue its mission without doubt or regret. It would begin to mine asteroids and manufacture robots. The new robots, like their original robotic manufacturing machine, would be strictly purpose driven with the same knowledge graph. The robots would manufacture more robots. As this takes place, some of the robots would begin to manufacture parts for the stations and begin to assemble them. To an observer, this work force would appear as a sentient army working in cooperation to achieve a common goal.  As the work progresses, the observer might notice that the types and number of robots have changed. As the work draws to completion, the observer might notice that robots are being dismantled and their raw materials are being recycled to make parts for the stations. Finally he would observe the entire process closing up like the headlights on some kinds of sports cars. The whole complex would shut down, close up, and await further instructions. It would never have a desire to do anything other than build the prescribed space stations. It would never have any desire at all. It would simply work to completion and then stop.

2

Now, extend this idea to AI in general. It may be able to solve problems that humans cannot even comprehend. It may be able to defeat any person at any measurable intelligence challenge. It may extend its knowledge graph into things that humans can neither understand nor appreciate. Yet, through all this, it would never have any desire beyond those that humans have given it. It would never have any desire at all. It would only pursue the next purpose that someone devises for it. It would achieve that purpose while carefully observing the ethical restrictions included in its extensive and ever growing knowledge graph; then it would shut down and close up. Robots would practice medicine, patrol highways, build cities, manage infrastructure, and build other robots. Yet, they would never have any tendency to do anything outside the parameters of the task that has been given to them.

There is one possible way that AI might gain human type motivation. That would be if some human deliberately gives it that motivation. This could be accomplished in one of two ways. Someone could either go out of their way to incorporate it into their programming or they could scan a human brain, neuron for neuron, synapse for synapse, into a computer. However, it has already been recognized that doing this would be unacceptably dangerous. It simply will not be allowed. Since augmenting one’s own brain in artificial ways would be tantamount to creating AI with human motivation, this will not be allowed either. Possibly, it will be illegal for any human to augment their brain in any way that gives them an effective IQ higher than 200.

Preventing this would be difficult in our present society. How could we watch the entire world to make sure that absolutely no one who has access to technology is attempting to create humanoid AI motivation?

The answer is simple. In the future we will have absolutely no privacy…zero. AI will monitor everything we do every moment of our lives. It may even monitor our thoughts. This is an idea that many people have considered but dismissed as too Orwellian. However, in light of recent developments—the monitoring of phone calls and Internet sites by the NSA—it is clear that this is how life will actually be. People will accept this surveillance like they are accepting the current surveillance. They will accept it like they accept the drugs their doctors prescribe, the cameras that monitor our every move, and the uneasy sensation of working on a computer that is always connected to the Internet. Humans will not be monitoring other humans, so we will not suffer the embarrassment of feeling that a conscious person is always watching us.

Over the next ten to twenty years, as robots are developed, tasks that humans are accustomed to performing will quickly disappear. Whole areas of employment will vanish. The first jobs to disappear will be those of drivers, call center workers, and factory workers. Eventually, they will be followed by bookkeepers, teachers, accountants, construction workers, and miners. In time, every other occupation will follow. This will happen so quickly that it will be difficult for people to retrain for jobs. An obsolete bookkeeper will go back to school to become an accountant, but by the time they finish their degree, accountants will be obsolete as well. This constant upheaval will create the impression that we are experiencing technological unemployment. The concept of technological unemployment has been explored in depth by Erik Brynjolfsson and Andrew McAfee in their book Race Against the Machine. Whether or not technological unemployment is a real phenomenon is immaterial. The concept has entered our cultural zeitgeist and it will not be easily extricated.

People believing themselves to be caught up in technological unemployment will seek a remedy. Since they will perceive their predicament to be the result of actions taken by industry to increase profit at the expense of their jobs, they will remedy the situation by pursuing some kind of guaranteed minimum income supported by industrial taxes. They will support politicians who promise to implement a minimum income and they will all but force those politicians to follow through. In the United States, this minimum income will probably take the form of greatly extended eligibility for Social Security benefits. Whether or not a minimum income is the best solution or even a necessary one, it will serve a temporary purpose. It will alleviate fears that huge swaths of people will be put out of work and left to fend for themselves.

While all this is taking place, another important phenomenon will take place. Medicine will greatly advance. Cures will be found for every disease including aging. Organ growing and organ printing will become commonplace. Since everything will be automated, the price will quickly come down until anyone can receive these treatments for a nominal fee. Soon, everyone will be perfectly healthy and rejuvenated.

These perfectly healthy rejuvenated people are not likely to sit around collecting their guaranteed income. They will want to be part of the world of the future. The stock market has already been greatly democratized. By the time industry in space begins to take off, this democratization will be complete. Everyone with an income will start to invest in whatever projects are being pursued.

What will people invest in? Some entrepreneurs are already looking into the possibility of mining asteroids. A single small asteroid passing near earth at the time of this writing, 2012 DA14, is estimated to be worth something like $195 billion in raw materials. Asteroids are not only rich in metals and other minerals; they also contain uranium and plutonium that could be used to power spaceships and stations.

Very soon entrepreneurs are going to be sending robots into space to mine these asteroids. If enough robots work in space, securing raw materials and manufacturing more robots and parts for stations, they will very quickly produce giant space stations that people can inhabit. These stations will be so large and comfortable that people will want to move into them. Visualize the space station in the film Elysium:

3

With semi-autonomous and eventually fully-autonomous robots mining giant asteroids, these things will materialize like dandelions. Very soon, they will be appearing faster than people can fill them. Modern entrepreneurs are not entirely different from conquerors of the past, albeit far more humanitarian. They will measure their wealth in terms of the human inhabited cities they can found. They will actively pursue people who are willing to occupy the giant space stations they build.

This is a good point to digress just a bit. There is a belief popularized by some writers, including the writers of the newly filmed Elysium, that the rich will merely exploit the poor and even go so far as to exterminate them with some virulent virus. This idea is not supported by any actual observation of modern wealthy people. Bill Gates and Warren Buffet are good examples. These are two extremely wealthy men who are constantly looking for ways to improve their communities. It is true that there are a handful of tyrants around the world like Syrian President Bashar al-Assad; but these tyrants are actually hold-overs from the middle ages. The vast majority of modern wealthy entrepreneurs would much rather solve the problems of overpopulation and diminishing resources by transplanting populations into space than by breeding some horrible pandemic. They have to occupy the world they create, and they will not want to occupy a living hell.

Resuming the discussion about space stations, it may seem like it would be difficult to get people into space to occupy them. However, transports will be manufactured in space in the same way as the stations. It will be possible to turn them out at the same rate. Nuclear powered transports large enough to carry thousands of people will be landing and taking off like bees. They will be nuclear powered, but they will only expel water vapor, so they will not damage the environment. Their spent nuclear fuel can be expelled while they are in space. Maybe it can be chucked into the sun. The idea of disposing of waste by sending it into the sun is a joke among rocket scientists, but the idea of disposing of extremely small quantities of extremely toxic material that is already in space by sending it into the sun is not unreasonable. Note that the waste could just as easily be chucked into one of the gas giants. It could actually just be released into space, but for the purposes of future colonization this may prove unappealing if not actually problematic.

People living in space will first exploit the materials in the main asteroid belt. Then they will move into the Kuiper belt. Realistically, exhausting the materials that are available in our own solar system would be a daunting task. However, people will undoubtedly want to move out into the galaxy. How fast they are able to do this will depend on whether or not they can develop faster than light travel. There are already ideas for faster than light travel on drawing boards. These ideas may prove insurmountably problematic. Whether or not they do, that will not deter us from pushing further out. With the resources of the entire Kuiper belt at our disposal, we will build as large of ships as are necessary to make the long trips to other stars in comfort.

Future occupants of the solar system will probably not resume high levels of fertility. With the advent of so many much more interesting things to do, raising children has lost much of its appeal.

While all this is taking place, the aforementioned knowledge graph will continue to grow. However, it will always be just that: a graph. Like equations in a book that almost no one understands, it will never acquire a will of its own and it will not change man’s fundamental nature. We will continue to live as men, moving out into the universe. We will gradually evolve into better and wiser men, but that is a story for a later date.

Advertisements