In transhumanism circles you’ll often hear talk about something called the “Singularity”. The way people talk about it, you’d think it was a religious term like “Ragnarok” or “the Rapture”. Actually, maybe those aren’t too far off from the sentiment around the Singularity, but the ideas and concepts behind it are not quite so mystical.
The Singularity, or more specifically “technological singularity”, is a proposed future event whereby the pace of technological progress will reach such a speed that the very essence of humanity will be transformed into something transcendent.
The idea that led to the Singularity as a concept grew from the observation that technological progress is not linear, but rather is best described by an exponential curve. For example, the technological progress we’ve experienced in the last 100 years is far more vast than that of the 1,000 years before it. The progress in the last 10 years is greater than the 20 before it, and so on.
Moored or Adrift?
When digital computing systems became a thing, especially after silicon microchips were invented, it became apparent that the performance of these devices were increasing faster than any other technology we’d ever invented. Gordon Moore, who is a co-founder of Intel, made an observation in 1965 that the number of transistors per square inch in microprocessors was doubling every 12 months. That’s an incredible pace of improvement and, apart from slowing down to every 18 months, “Moore’s Law” has held true ever since.
If you project this growth into the future, under the assumption that it will keep happening, you reach a point where computer systems are so powerful that there’s really nothing they can’t achieve. Not only will they be capable of matching human intelligence, but they will quickly move beyond it.
The key moment comes when artificial intelligence takes over it’s own development. For most of the history of computing it has been human researchers, engineers, and scientists that have driven this development, but when a given generation of computer is smart enough to design its own successor, then all bets are off in terms of speed.
We’re already seeing some of this with modern electronics and circuit designs. Evolutionary computer algorithms are being used to design the most efficient circuitry possible by automatically simulating millions of permutations and then testing their “fitness” in a computer simulation. In a very short span of time, the computer can perform the sort of trial and error that could take years or decades for human designers.
A Brain Explosion
The Singularity is when we see the birth of that “super intelligence” – an entity so smart that it will also quickly become immensely powerful. It will solve problems of physics and chemistry we’ve never even thought of. Technical achievements such as interstellar travel, terraforming, and mind-uploading may only be possible (or possible in the near future) with the aid of such a super intelligence. It’s no wonder that many people seem to think that the Singularity is a religious movement, since such a superintelligence would certainly qualify as a “god” from a certain point of view.
If the Singularity is going to happen, when will it happen? There are many guesses, some more educated than others, but the general average seems to center around the year 2040, which is not that far away from when I’m writing this right now!
Not So Fast
One of the key failure points in the premise of the Singularity is the assumption that the increase of computing power will continue indefinitely. That’s a big “What if?” though. After all, the past does not predict the future. It just provides data to model the future. We simply don’t know if the increases in computing power will keep pace.
It’s clear that silicon processors are not going to be the technology to take us all the way there. We’re close to hitting the hard limits of physics when it comes to how small you can make transistors. There comes a point where the circuitry is so small that quantum effects on electrons become important; silicon is not up to handling any more shrinking of component sizes.
Many research teams have been working hard to figure out what material or technique will take over from silicon. The lead candidate for this is carbon nanotubes, although it must be said that this is a material that suffers from incredible hype. Still, IBM announced some very important breakthroughs in using them as a replacement for silicon semiconductors.
Hopefully, this means that carbon nanotube CPUs will take over smoothly from silicon when silicon does hit the wall. The jump may even be much bigger than we anticipate.
Back in 2010 IBM demonstrated that graphene transistors, which are also made from carbon, were more than twice as fast as the same design in silicon. The silicon system could hit 40 Ghz, versus the 100 Ghz figure for graphene. That would provide plenty of breathing room to extend or even accelerate Moore’s Law for a while yet. IBM’s team estimated that if their transistor was brought down to nanoscale, as it would be in a CPU, it could hit 1 terahertz in speed.
But perhaps even the wonder-material that is graphene will hit a roadblock eventually and will likely not be enough to bring about a Singularity-like super intelligence. Luckily, there are many other radical computing technologies on the horizon. It’s almost inevitable that future computing will consist of a mix of these different approaches to computation and artificial intelligence.
Quantum computers, for example, can solve some types of problems that digital computers simply can’t. They may also hold the key to some aspects of consciousness, although that’s still speculative. DNA computers are yet another type of radical computer that could make up a part of a future super-intelligence; computing with the actual hyper-complex chemistry of DNA. Finally, we may see computers that use light instead of electricity. The photonic computer is a theoretical computing device that is still only an idea, although some computing components have photonic equivalents that could work.
What the Singularity Means for Us
The Singularity would be an apocalypse or paradise, depending on who you ask. For non-transhumanists it may be clearly apocalyptic to humans, and that would be factually correct. Humans in their current form would most likely cease to exist after the Singularity. What would be left are post-humans, in whatever form they would take.
From a transhumanist perspective the Singularity could be highly-desirable. In a way it represents the transhumanist ideal. It mean intelligence given almost unlimited reign and an existence beyond the confines of being human.