In 1993 the science fiction author Vernor Vinge wrote a paper in which he predicted a coming event which would radically change life as we know it. “I argue in this paper”, he wrote, “that we are on the edge of change comparable to the rise of human life on Earth. The precise cause of this change is the imminent creation by technology of entities with greater than human intelligence”. Vinge coined a name for this change. He called it the ‘Technological Singularity’.
What I want to explore in this essay is the possibility that a singularity is not a unique event but one which has happened more than once. I believe that Vinge’s own words lead us to suppose that reality has gone through profound shifts in possibility before. The key sentence is as follows:
“This change will be a throwing away of all previous rules…developments that before might only happen in “a million years” (if ever) will likely happen in the next century”.
In other words, a key aspect of a singularity is that it leads to a dramatic change in perceptions of time, or rather, a dramatic compression of possibility, such the the wildly implausible becomes likely. With that in mind I think we can see in the past history of our universe at least three events which qualify as Singularities.
SINGULARITY NUMBER ONE: THE BIG BANG
Before science fiction writers speculated on the possibility of technology bringing about such a dramatic change that it imposed a ‘singularity’ on our future through which we could not peer and see clearly what was to come, cosmologists looked to the dim and distant past and traced the evolution of the universe itself until they reached a point where our understanding of physics can take us no further. They called this point where the state of existence is shrouded in utter mystery a ‘singularity’ and no doubt those speculators of the future (Vinge was not the first person to use the phrase in the context of future technological change) borrowed the phrase from the cosmologists.
Today, by far the most popular theory of the universe’s origins is ‘The Big Bang’, which points to evidence that our universe is expanding and argues that, if this is so, as we go back in time the universe must have been smaller until a moment is reached where all of creation was compressed into a speck of infinitesimal size. This naturally leads to the question, ‘what happened before?’, to which the answer seems to be ‘there was no ‘before’. The Big Bang marks the moment time and space began, so asking what came before the Big Bang is as nonsensical as asking ‘what’s north of the North Pole”? It follows from this that it is meaningless to wonder how long that mysterious pre-Big Bang reality lasted, for without time a nanosecond is no different to an eternity. What bigger difference in perceptions of time can there be than the transition from timelessness to change that can be measured?
SINGULARITY NUMBER TWO: THE ORIGIN OF LIFE
Vernor Vinge saw the Technological Singularity as an event which could compress our expectations of what is possible in a given time-frame, making ‘only in a million years’ events happen within a century, if not sooner. If we look to our past, we can see another event which dramatically speeded up possibility, and that event was the transition from single-step selection to cumulative selection.
Richard Dawkins illustrated single-step selection by selecting a phrase from Shakespeare’s ‘Hamlet’ (‘METHINKS IT IS LIKE A WEASEL’) and asked how long we should expect to wait for a monkey, randomly bashing away at a special word processor with only 27 characters (each letter of the alphabet-capital only-plus a ‘space’ as the 28th character) and which allows only exactly 28 bashes per go, which happens to be the same amount of characters in that phrase, if we include a space as a ‘character’.
The chances of a monkey happening to type ‘M’ as the first letter is one in 27. After all, there are 27 other possible characters that the primate could happen to bash. In order to get the first two letters the monkey must beat odds of 1/27 times by 1/27, which gives us odds of 1/ 729. In order to randomly type the entire sentence with no spelling errors and spaces all in the correct place, the monkey must beat odds of about 1 in 10,000 million, million, million, million, million, million. As you can imagine, then, you would likely have to wait a very, very long time for a monkey to bash out that precise phrase on Dawkins’s special keyboard.
And yet these odds are quite good compared to the odds of a haemoglobin molecule happening to assemble itself from the random recombinations of amino acids from which it is made. The haemoglobin molecule consists of four chains of amino acids, there are 146 amino acids per chain, and in living things we commonly find 20 different kinds of amino acids. Another science fiction writer- Isaac Asimov, calculated the number of possible ways of arranging 20 kinds of things in chains 146 links long and came up with the ‘haemoglobin number’, which is (more or less) a one with one hundred and ninety noughts after it. Compare that to our ‘METHINKS’ odds in which the monkey ‘only’ had to beat odds of 1/10^40 or 1 followed by forty zeros. And, of course, one haemoglobin molecule makes up only a tiny fraction of the complexity of a living organism. If it were left up to random chance, we would have to wait far longer than the life of the universe itself for life to emerge.
Since life evidently has emerged we must conclude that a Vinge-style ‘possibility compression’ must have occurred at some point in the past, and we know exactly what that event was (although we are still in the dark as to what exact form it took). That event was the transition from single step selection aka random chance to ‘cumulative selection’.
The difference between these two is that, whereas single-step selection has no memory whatsoever of the past, where cumulative selection is concerned the results of one process is fed into subsequent processes. To illustrate the power of cumulative selection, Dawkins designed a computer program that, like that monkey, randomly a random sequences of 28 letters. It would then duplicate that phrase but with a certain chance that ‘copying error’ would alter the phrase. The computer would then examine all those ‘offspring’ phrases, and reproduce whichever phrase most closely resembled ‘METHINKS IT IS LIKE A WEASEL’.
First of all, the program typed the following:
WDLMNT DTJBKWIRZREZLMQCO P
Pretty much the kind of thing you would expect a monkey to produce were it let loose on a word processor. After ten generations and selecting of ‘phrase closest to METHINKS IT IS LIKE A WEASEL’ the program had managed to produce:
MDLSMNS ITJISWHRZREZ MECS P
Still hardly a recognisable word, let alone Shakespearian in its quality.
By the time the 30 generations had been bred and selected, a resemblance to the target phrase had become undeniable:
METHINGS IT ISWLIKE B WECSEL
And within 43 generations cumulative selection had produced the exact quotation.
How long did it take for the computer to evolve five word quotation from ‘Hamlet’? About eleven seconds. Compare that to how long we would expect to wait if we relied only on random chance: About a million, million, million, million, million years.
Now, there is one important difference between Dawkins’s evolutionary program and natural selection, and it is this: That program was given a definite target in that it had to search through strings of 28 characters and select the one which most resembled, however slightly, the phrase ‘METHINKS IT IS LIKE A WEASEL’. Natural selection, on the other hand, is not heading toward any definite future goal. But still, the experiment Dawkins run does give us some inkling of the power of cumulative selection to dramatically speed up the likelihood of something as improbable as the complexity of life as we know it today. Paraphrasing Vinge, we might say that ‘events which would otherwise take about a million, million, million, million, million years to happen, can actually happen between eleven seconds and one hour’.
SINGULARITY NUMBER THREE: THE GREAT LEAP FORWARD
The theory of evolution by natural selection tells us that human beings are just one more species belonging to a great family tree comprised of all living things that exist, or have ever existed. But are human beings really just another animal, no more remarkable than any other creature or plant? Or is there a good reason to pick human beings out for being special in some way?
I think the latter is true, and the special reason is as follows: Human beings, unique among life on Earth, enabled a new kind of evolutionary process. As Dawkins wrote, “There is an evolution-like process, orders of magnitude faster than biological evolution…This is variously called cultural evolution, exosomatic evolution, or technological evolution”. Whereas all other forms of life on Earth can only adapt at the speed of natural selection, human beings have the imagination, the communicative capability, and the dexterity to reshape materials around them to produce useful designs intended to suit some purpose. We don’t have to wait for natural selection to adapt us for operating under water, we can develop snorkels, aqualungs, submarines and other forms of aquatic technology. And while it took billions of years for natural selection to produce flying animals like birds, it took only a couple of million years for human beings to fly to the Moon.
Our hominid ancestors were not always so rapid in their technological development. If we look back more than forty thousand years we find man-made artefacts that had hardly changed for a million years. Generation upon generation upon generation of humans produced the same kind of flint knife that their ancestors relied on, plus a few other tools. As for paintings and carvings and figurines, they produced none.
Our distant ancestors were no different to any other animal. Humans are not the only animals that make use of tools. Other primates have been observed using blades of grass to ‘fish’ ants out of holes with; thrushes use a stone as an ‘anvil’, bashing snails against it so as to break its shell and get at the soft meat inside; beavers fell trees in order to construct dams, the list goes on. But none of those animals have anything like technology, which is an ever-accumulating family of tools and techniques solving more and more problems. No beaver ever figured out how to add hydroelectric power to its dam, no thrush ever learned to add its snail meat to a recipe combining that ingredient with others to produce a tastier dish. Similarly, it seems our 40,000 year old+ ancestors never figured out that they could make dramatic improvements to their tools and that there was an almost infinite range of possible tools and techniques that were waiting to be fashioned from the resources around them.
But then, something happened, and Jared Diamond called this event ‘The Great Leap Forward’. As I said before, prior to this Leap, the tools our ancestors used hardly changed for a million years, but after it we find paintings, carvings, musical instruments, and the beginnings of true technological capability that would result in, among other wondrous inventions, the iPad2 and app on which I am writing this very essay a mere 40,000 years or so later.
As Matt Ridley wrote in ‘The Rational Optimist’, the human race has “surrounded itself with peculiar, non-random arrangements of atoms called technologies, which it invents, reinvents and discards almost continuously. This is not true for other creatures…they do not ‘raise their standard of living’, or experience ‘economic growth’…they do not experience agricultural, urban, commercial, industrial, and information revolutions”.
This leads one to ask what it is about the human species that enabled it to trigger this paradigm shift to technological evolution. Some authorities think it has something to do with language. Perhaps not the evolution of language itself (linguists like Stephen Pinker believe language to be older than the Leap) but rather (as Dawkins speculated) “a new trick of grammar, such as the conditional clause, which, at a stroke, would have enabled ‘what if’ imagination to flower”. You can understand why people look to language, or some adaptation of the ability to communicate linguistically, for the triggering of the Great Leap Forward, because technology is an inherently collaborative process. The idea of the lone genius who gets a great idea from out of nowhere is pretty much false. Instead, inventors take materials, tools, techniques and ideas that already exist and put them together in a new way to achieve a new result. They rely, in other words, on the work that has already been accomplished. Furthermore, they pretty much always rely, either directly or indirectly, on support from other people in building whatever they are making. This fact is illustrated in a famous essay called ‘I, Pencil’, in which you are challenged to make the kind of pencil you can buy in a shop from scratch. Of course, you could snap off a twig and use it to etch markings into soft Earth, but could you make a pencil with a graphite tip and a little rubber on the opposite end, held in place by a strip of metal? In order to do that you would have to know how to mine that graphite and metal, produce that rubber, and how to get that graphite inside a hollow tube of wood. And, of course, you could not rely on anyone else’s tools to do this work, you would have to build it all entirely from scratch.
Simply put, this is an impossible task for one person to do. All of us rely on work that was done almost entirely by other people. Technology is very much dependent on specialisation and exchange, on people collaborating with one another and relying on an accumulating, evolving record of knowledge. Such a capability could never have arisen had each individual only had its own mind to rely on, and no way of communicating ideas from one mind to another and across generations. Vinge himself actually used the arrival of the human species as an example of a Singularity-like change. Recall that he wrote:
“We humans have the ability to internalize the world and conduct “what if’s” in our heads; we can solve many problems thousands of times faster than natural selection”.
So, there we are: My three examples of past events which had such a dramatic effect on time, on what is possible, that they deserve to be thought of as ‘Singularities’. Makes you wonder how different the future will be if Vinge’s ‘technological singularity’ actually happens, doesn’t it?
The Blind Watchmaker, and The Ancestor’s Tale” by Richard Dawkins.
“Rational Optimist” by Matt Ridley.
“The Technological Singularity” by Vernor Vinge.
“I, Pencil” by Leonard E. Read.