Words

Introduction
The “Technological Singularity” defines a subset of the possible (and in the opinion of many futurologists, such as myself, very likely) futures that could be in store for the human race. The Wikipedia article on the Singularity has this to say about the term itself:

“…the ever accelerating progress of technology and changes in the mode of human life, which gives the appearance of approaching some essential singularity in the history of the race beyond which human affairs, as we know them, could not continue.” — Stanislaw Ulam, May 1958, referring to a conversation with John von Neumann

Picking Up Speed
Over the next few decades, the thinking goes, our world will be witness to the most astonishing leaps in technological prowess in all of human history. Nanotechnology, once mature, will be capable of repairing biological damage at the DNA level itself. (Getting it cheap enough to deploy on a scale large enough to heal a whole human body is another matter, but I am confident it will happen, just as I am confident that one-terabyte hard drives will be considered old-hat in five years.) Artificial intelligences may match or even excede human intelligence within the same time frame. Technology, in general, will continue to improve at an ever-increasing rate. What will life be like in 2030? Hard to say, but I’ll say one thing: Look at the rate of technological improvement between 1980 and 2005, and know that all of that incredible progress is but a shade of what we will do next.

Accelerating Returns
Alright, let’s get down to brass tacks. One of the strongest predictors for the Singularity is Kurzweil’s Law of Accelerating Returns. In short, technology develops on an exponential curve, rather than the somewhat straight line it may appear to develop on. It is very easy to look at what has changed in technology over the last year and project what will happen over the next, and to see something like a straight line between a year ago and a year from now. However, if you look at technology over a longer timescale, you will see that it is, in fact, on an exponential curve. Every comptuer scientist has heard of Moore’s Law, which states that transistor density in semiconductors is able to double about once every 18 months. Quoting the Moore’s Law article linked above:

As of Q4 2004, current PC processors are fabricated at the 130 nm and 90 nm levels, with 65 nm chips being announced by the end of 2005. A decade ago, chips were built at a 500 nm level. Companies are working on using nanotechnology to solve the complex engineering problems involved in producing chips at the 45 nm, 30 nm, and even smaller levels—a process that will postpone the industry meeting the limits of Moore’s Law.

This is not really a prediction that can be generalized to all technology - just semiconductors. However, it does form the basis of Kurzweil’s Law of Accelerating Returns, which states that the rate of technological progress - overall - doubles once every decade.

According to Kurzweil, you can assume that a hundred years of technological progress - at the current rate of development - will actually take just 25 years, since we continue to learn things that allow us to do things faster, better, more reliably, more cheaply, etc.

For Example…
Let us say that in order to design Deep Thought using today’s computers and other technology, design and research methodology, and so forth - never using any newer technology, etc. - it would take 100 years. If we periodically upgrade the computers we are using to design Deep Thought, and migrate to newer and better ways of doing research and development on the project, it will actually wind up taking just 25 years to design Deep Thought. This is because computers will continue to increase in power over time, and new efficiencies will be achieved with better R&D processes.

Artificial Intelligence
The physical basis of the human mind - that is, the majority of the brain - is a neural net. It is a learning computer. It starts out almost completely blank, with only a number of sensory inputs and neuromotor outputs and some built-in instincts to cause it to do anything more than sit in a catatonic state. As the owner of the neural net progresses through life, its internal state changes in ways that are informed by the senses and instincts innate to human existence. I believe that strong AI - that is, AI capable of tackling general problems around or above the level of a human, rather than just a single problem domain, such as playing chess - will eventually come together. This will arise from advances in neural net research, knowledge bases such as Cyc, genetic algorithms, our increasing understanding of the human brain, and other areas.

The thought process here is that strong AI may play a pivotal role in the Singularity. Let’s say we design an AI that is as intelligent as the human mind, and then set it to work designing more and more intelligent AIs. One day we have an AI that is a thousand times more intelligent than a human. What do we do with that AI? How do we control it? How do we harness its power? What happens when there are two of them, and then six, and then a thousand, and then a million? Tricky, isn’t it.

Alright, what’s the score here? What’s next?
The term “Singularity” is meant to imply that as we get closer to the actual event itself, any assumptions we have about what the future will hold will become increasingly unreliable. We may have a general idea, in much the same way that people in 1980 had some inkling that computers weren’t just a passing fancy. However, the specifics become rather fuzzy; there were few in 1980 who had even heard of the Internet (as, in this case, a use for computers), much less had any idea that it would become a pervasive element in the human experience, as television and radio did in the twentieth century.

To use the same analogy, we can look at current technology and how it integrates into society, and make some projections about what the future may hold. We probably won’t be dead-on with our projections, but it is arguably worthwhile to at least make some attempt at figuring out what the world might be like in the future. I think the next 25 will be - as Zaphod Beeblebrox would say - Amazingly amazing.

In my next Big Post About Fantastic Things, I will write about… the future!