Category Archives: Economics

Intelligence – Very, Very Fast Intelligence

Speed superintelligence is quite simply the ability to do what a human can do, but much faster.

Through training, we are able to get faster by optimizing our neural circuitry. The result is that our reactions become faster, allowing our performance to increase. This is most evident when we pair a well-trained athlete with an athlete of a relatively low skill level. To the beginner, everything seems to be progressing at a breakneck pace, barely able to keep up. To the well-trained expert, time seems to be moving at a slow, almost relaxed pace. We often say that the pro makes it all seem effortless.

This optimization through training yields an improvement by a factor of two or three. In a machine, we will realize improvements by a factor of hundreds or thousands. In specific problem areas, we already have machines which are faster than humans at a factor approaching one million times.

Imagine a scenario where you are competing with a machine for a job at a bank. It might be a position processing loans or mortgages, or perhaps managing a portfolio of investments. Both of these jobs already leverage computers for accessing and processing in formation. But today we still rely on human loan officers and portfolio managers to make decisions about which loans or mortgages to approve, which stocks to buy, and when to sell them. But what if we had enough confidence to let the computers make all of these decisions on their own, without supervision? How much more quickly could these computers process loans or make buying and selling decisions on investments? These automated financial agents could easily make decisions and process transactions at a rate that would allow them to take the place of tens or even hundreds of their human counterparts. In fact, they are used extensively in the financial sector today and continue to become more pervasive. Automated trading has been used to accelerate trades for years and has now become available even to the individual investor.

From the perspective of speed and efficiency, machines clearly are more intelligent than humans already. They have become so good at what they do that they make their task seem effortless.

How the Acceleration of Technology Will Allow Computers to Take Over the World

This idea that computers will advance at a rate which will allow them to surpass the capabilities of humans is not a new one. One of the thought leaders in this area, Ray Kurzweil, has long been a champion, if not the originator, of the notion that technological advances increase at an exponential rate. This is not simply an observation that technology is advancing faster and faster, but that because of the nature of the advances, the improvements delivered by technology build on one another. Kurzweil has laid out in detail what he calls the law of accelerating returns (LOAR) in his first three books, most notably The Singularity is Near. He explains that the hierarchical nature of technology is what enables this exponential growth. The evolution of technology occurs in increasing levels of abstraction resulting in exponential complexity and as a result capability. A set of technological advancements is built upon to form a new and more complex innovation with far more impact than the individual components themselves have. Consider the evolution of electronics as an example. Just as we were completing development of the world’s first modern computer, the ENIAC, three physicists at Bell Labs were inventing the transistor, the fundamental component of all modern day electronics. The transistor allowed us to build electronics such as radios, calculators, and computers at a fraction of the size and cost of the same devices formerly built using vacuum tubes. Two years later, the integrated circuit was invented, putting five transistors on a single chip. Fifty years on the IC or microchip technology had advanced to the point where we could put millions of transistors and other electronic components on a chip the size of a fingernail. Within the space of fifty years we went from building computers with tubes and wires which took up an entire room and could only perform relatively simple calculations to machines small enough to carry around with us every day and could outperform even the most powerful computers of the twentieth century.

In their book The Second Machine Age Erik Brynjolfsson and Andrew McAfee discuss many of these same issues primarily from a socio-economic perspective, stating “Rapid and accelerating digitization is likely to bring economic rather than environmental disruption, stemming from the fact that as computers get more powerful, companies have less need for some kinds of workers.” This is not a case where we gradually lose ground to automation, the advances in machines are accelerating and have been ever since the introduction of the transistor. The impact was initially not evident and has been underestimated for some time, but just as the industrial revolution enabled vastly increased productivity within a few decades this second revolution will enable increased productivity in a different way. The industrial revolution addressed physical limitations of humans such as speed, strength, and consistency. This current advance in technology will address the non-physical, the power of the human mind. Computers can already do more than the human mind in very many areas such as processing large amounts of data and performing lengthy calculations. We are now starting to see the use of computers to predict our wants and needs when we visit a website or to tell us the fastest way to drive to the mall. We carry devices which tell us if we are getting enough exercise or eating too much. This next generation of computing devices will benefit us greatly, assuming much of the burden of everyday life and doing a better job. But the benefit will not come without sacrifice. The emergence of assembly lines allowed for faster and cheaper production but at the cost of the extinction of certain jobs. In the same way, many of the tasks which we rely on humans for will be performed faster, better, and cheaper by our silicon assistants. Just how far will this wave of succession extend? Will we ultimately find ourselves at the mercy of a society of robots, relying on them for every one of our needs to survive? The story has been the subject of countless science fiction such as “I, Robot” and “The Matrix”. Much of the future is still unknown, we haven’t lost control just yet. But the shape of things to come is evident and undeniable. As Brynjolfsson and McAfee summarize, “In short, we’re at an inflection point— a point where the curve starts to bend a lot— because of computers. We are entering a second machine age.”

Welcome to Computers Can Do Anything

We all like to think that there is something superior about humans. We don’t just think we are better, but that we possess some quintessential element that machines do not. The basic premise of this site is that computers can do anything you can do and do it better. There are many examples of things that computers do not do well today. There are even some things that computers can not do at all. Each year we see improvements in the capabilities of our silicon friends. And each year machines, computers, and other electronic devices take on more and more of our everyday tasks, from balancing our checkbook, to flying our planes. And although believe that there are some things that a machine will never be able to do, there is no limit to the capabilities of the thinking machine. In fact, there will come a time when machines will be superior to us in every way.