Transhumanism Singularity index page The Singularity Nature, Humanity
and AI Harmony
Books Links Search entire Site

Musings on the Singularity

When I was a kid, I used to eagerly devour science fiction.  The works of all the old masters - Asimov, Clarke, Heinlen, Bradbury, Van Vogt, Herbet, various short stories, and of course the television staples Star Trek (the original series naturally), Dr Who, and Lost in Space. 

As a teenager I used to draw my own science fiction comics, design blueprints for space ships and ray guns, imagine aliens and robots and other planets and interstellar civilisations.  When I was a little older and Star Wars hit the big screen, I marveled at the special effects of gigantic spaceships and alien races.  (another amazing thing was that the bad guys never seemed to be able to shoot straight - marksmanship obviously not being a part of stormtrooper boot-camp training).  I thought Star Wars presented a pretty passable picture of a totalitarian future galactic empire, conveniently ignoring the opening lines "in a far distant galaxy a long time ago..."  It's the future that I was (and am) interested in, not a fairy tale about the past.

Unfortunately, if what mathematician Vernor Vinge and the transhumanists are saying has any truth to it (and I believe it does), I got it wrong.  Quite wrong.  And so did Asimov, Van Vogt, Gene Rodenberry, George Lucas, and all the rest.

Quiet simply, what Transhumanists and Extropians suggest, is that within a very short time, possibly within the lifetime of most of you readers out there, history is going to end.  Yes, that's right.  And we're not talking about some Fundamentalist Christian Rapture or New Age Cosmic Photon Beam here.  This is all hard science and technology.

The Australian science fiction writer Damien Broderick calls it "the Spike."  Because when this phenomenon is charted on a piece of graph paper it looks like a  spike.  The preferred term among those who are anticipating this phenomenon is "The Singularity."  The term was introduced in this context by Vernor Vinge, a mathematician and computer scientist at San Diego University (see Vinge's paper Technological Singularity)  Vinge observed (although he was certainly not the first to do so) that technological progress tends to increase in an exponential rather than a linear manner.  What this means is that not only does knowledge and technology progress, but it progresses at an ever-accelerating rate.

Consider the fact that it took a million or so years for the first crude stone-chipped tools to develop into somewhat more sophisticated forms, another couple of hundred of thousand years for the next stage in stone age technology, then several tens of thousands of years after that for agriculture and the first urban settlements, another few thousand to get the wheel, writing, and bronze tools. You get the idea.  So now we live in a "future shock" world where fads and fashions change constantly and the most advanced personal computers become obsolete in two years.

What Vinge did was to ask: if this keeps up, then what?

What happens, he suggests, is that you get to a stage beyond which it is literally impossible to predict what will happen.  Vinge termed this limit "the Singularity".  In mathematics and physics a singularity is a point where infinities enter the equation and analysis breaks down.   A black hole is one example of a singularity; any object sucked into its maw would find the laws of physics cease to apply.  The first moments of the "Big Bang" constitute another example.

In the sense that Vinge uses the term, the Singularity is not a physical thing in space-time, but a sort of barrier to prediction.  We can look at current trends and get some idea how the world will be like in, say, 5 or 10 years time.  But accelerating trends in computer sciences converge very soon after that to form a wall technological novelties making any further prediction impossible.

There is a phenomenon known as "Moore's Law" proposed in 1965 by Gordon E. Moore, one of the founders of Intel, the computer-chip manufacturing giant.  This states that the number of components on a for computer chip, and therefore its processing power, doubles every 18 months or so.  Almost a third of a century on, this trend is still holding up quite happily.  If it continues, we can expect a mainframe supercomputer equal in computational power to the human brain by 2010, and a desktop version by 2030.

What happens if you use one of these machines to simulate the workings of the human mind?  What happens if it "wakes up"; becomes self-conscious?

But remember, things don't stop there.  Progress is accelerating all the time.  As Vinge wryly observes "what do you build five months after that?  Or what does it build five months after that?"

Consider a self-aware Artificial Intelligence (AI) designing an even more intelligent and powerful AI, which in turn constructs an even more powerful AI...

Before we arrive at that apocalyptic event, Broderick in his book The Spike foresees a lot of other strange new things.  He runs through a few attempts at a sort of quasi-immortality.  There's cloning; that's already here (as illustrated by a now very famous sheep called Dolly).  Then there's cryonics - freezing your body, or just your head! - so that at some future point you can be revived and reconstructed.   Or chromosome mapping and genetic engineering in the hope of curing diseases like cancer and prolonging life, possibly indefinitely.

Then there's full-scale sensory-immersion Virtual Reality; a sort of wired body-suit perhaps with detailed visual, auditory and tactile feedback.

But nothing is quite so astonishing, or promises so much, as molecular nanotechnology, the prospect of building things atom and atom.

Nanotech was conceived in the 70s by Eric Drexler and the idea has now caught on, although a practical nano is still a long way away.  It is hoped that eventually (by which I mean in the next few decades - accelerating technological progress again) it will be possible to construct self-replicating machines (called "assemblers") the size of large molecules.  This assemblers will construct other nano-machines which in turn will build anything.  And I mean anything - fried eggs out of old tires, starships out of dirt, a million Mona Lisas (identical in every detail) out of the contents of the local rubbish tip.  They'll be able to do this because they will select their raw materials atom by atom.  It will be, so it's claimed, the end of the age of want and the beginning of the age of limitless abundance.

I actually have some problems with this scenario; not so much the technological side of it as the fact that vested interests are not likely to welcome a technology that renders our entire economic and manufacturing system obsolete.

Well, never mind.  Once there are super-intelligent artificial or machine-augmented minds running on quantum supercomputers (built up from nanotech constructed molecular circuits) it won't really matter anymore.

Of course, several objections can be raised.  First, how do we know that the curve really is a spike, that it won't level off once the boundary limits are reached.  (e.g. modern cars don't go any faster or have any more horsepower than cars in the '30s and '40s)

Some philosophers actually object that a computer or a computer program could ever become self-conscious, or that uploading is at all possible.  Consciousness, they would say, is a unique property of organic brains ("wetware").  To me, this seems like parochialism.  Why shouldn't consciousness be able to exist in other forms?  Many mystic teachings even state that consciousness exists everywhere in nature.

My personal feeling is that Man's reign on Earth is coming to an end.  Within half a century a new form of sentient intelligence will take over as the dominant intelligence on this planet, and in this little corner of the universe.

Kheper Home | Singularity Index | Future - the Lobby | Transhumanism Index | Topics | New or updated | Search

images not loading? | error messages? | broken links? | suggestions? | criticism?

contact me

Creative Commons License
Unless otherwise attributed or quoted, all text is licensed under a non commercial attribution Creative Commons License 1.0 and a 2.0.
content by M.Alan Kazlev
page uploaded 14 March 1999
page last modified 4 July 2005