theultimatecomment.com

a psychedelic cult working towards the secret of the universe

The singularity


The singularity (or technological singularity) is an insane event that some hypothesise will happen in The Future, when the rate of technological change become so fast it becomes effectively infinite.

The statistics show that technology improves at a surprisingly predictable rate. If you try to predict a fundamental breakthrough, like when will teleportation be invented, that is hard to say, but if you look at improvements on existing technology, that does tend to be predictable. And many technologies grow exponentially, not linearly. People tend to think linearly - we think because we had certain changes in the past ten years, so we'll have a similar number in the next ten. This is like saying that if computer speed doubled last year, it will increase by the same amount this year. But it won't; if it doubles every year, it will increase twice as much this year, four times as much next year, eight times as much the year after that - and in thirty years it will increase over a billion times. This is not merely a theory; there is considerable statistical evidence. In fact, in the past thirty years, computer speed has increased almost exactly a billionfold.

You might've heard of Moore's Law, which states that microchip speed per dollar grows exponentially. Less well-known is that there are hundreds of exponential technologies. The size of the Internet doubles every five years. The rate of computer processing has doubled since the days of vacuum tubes, into the days of transistors and smoothly into the age of microchips. When the speed of microchips hits a wall, there are other technologies waiting in the wings, like quantum computers or the 3-D processors currently being invented at Intel. Data storage per unit of space or unit of money doubles every 15 months. The bandwidth of wireless information-transfer is doubling every 30 months and has done so consistently since 1897. Technology is becoming five times smaller every ten years. The resolution of body-imaging devices is doubling every year and its speed is doubling every two years. The total amount of solar energy produced is doubling every two years. In genetics, the cost of sequencing DNA is halving every year, while simultaneously becoming faster and requiring smaller devices. This is all exponential growth, meaning that in the next few decades, they're gonna clock some insane increases, the sort that'll transform society.

This is called The Law of Accelerating Returns. As technology advances, the rate at which we can create new technology advances. If we are twice as advanced, we can make progress twice as fast. Therefore, as mankind becomes more and more advanced, the time taken to make technological advances becomes shorter and shorter, until it reaches a point where this time approximates zero. This point is the singularity.

So what does this mean for me?

It means you're living in interesting times, baby. The next 50 years will see more social change than the previous 400 years. The 1990s saw a revolution in information technology that transformed all our lives with mobile phones and the Internet and all that. But the current statistical trends show that there are at least three revolutions looming that have the potential to be equally revolutionary.

The first revolution will be genetics. The Human Genome Project was the first to successfully sequence human DNA. It took 13 years and \$300 million. Now there are companies that will sequence your DNA for \$4400 and it takes about a week. I prophesy that before 2010 is out, there will be a \$1000 DNA sequencing service. Some time before 2013, it will become standard practice for everyone to get their DNA sequenced, just as everyone now can get an eye test. This will transform medicine, especially preventative medicine, as everyone will know exactly what dicky genes they have and what to look out for. It's not much good to read DNA if you don't know the language, so we need to identify what genes do what. I'm not sure if anyone's statistically worked out the growth rate of identifying sequences of DNA, but it makes sense that as the technology becomes cheaper and faster, more people get their DNA sequenced, giving more data in which more ever more advanced computers can find patterns, leading to exponential growth of identifying genes. Our ability to write DNA is about 8 years behind our ability to read DNA. Before 2020 (just ten years away, man!) the genetics revolution will have come to full fruition, with desktop devices capable of reading and writing DNA at costs that make it accessible to everyone. We will be able to easily build, by 2025, anything that can theoretically be build out of proteins: stronger and harder wood, a superabundance of food, bacteria that generate electricity from poo. If you think Internet cat memes are weird now, wait til genetic engineering takes off. Let Andrew Hessel tell you all about it. I don't know if there will be genetic engineering for non-medical, human-enhancement reasons, but I don't see why not. If this increases human intelligence, then the rate of technological change ramps up even further. Unless some asshole has killed us all with smallpox by then.

I used to worry that technologies like these would create an elite of rich people who rule over those who can't afford it. But look at computers and mobile phones. Did we get a power-elite of mobile phone users, taking advantage of their advanced technology to dominate the lowly peasants unable gain such telepathic abilities? Do rich people have privileged access to the Internet, allowing them superior information to take advantage of the rest of humanity? Or did these technologies empower ordinary people and act as a democratizing force that decentralized power? The nature of exponential technology is to get really cheap really fast and this means that there is a period when only scientists, academics and the military have access to it (think of computers in the 1960s), but they soon spread to even the lowest reaches of society (like the Internet is now spreading to the poorest people in the poorest countries). Genetic engineering is starting to get out of university laboratories into attics, basements and garages, see DIYbio. As of 2010, eBay has DNA synthesizers for less than \$3000 and DNA sequencers for \$1000, well within the range of a dedicated hobbyist, and there is a directory of genetic sequences that is free and open-source. So both genetics and information technology have proven to be democratizing, decentralizing forces.

The next twenty or thirty years will also see a fabrication revolution, overlapping considerably with the genetics revolution, but lagging a little bit behind. We are exponentially moving towards a point where we have the same mastery over matter that we now have over information. That is to say, we will be able to mould matter according to any whim on our desktop. It is likely that manufacturing will become decentralized: rather than things being made in a factory and shipped out to people, the designs will be downloaded online and the physical object printed at home. This is already happening with RepLab, a self-replicating, open-source plant capable of manufacturing any mechanical or electronic design. But (again analogous to computers in the 1960s) the technology is too bulky and difficult to operate to really catch on. At the same time as these large manufacturing machines are improving, molecular nanotechnology is improving, moving towards a point where we will be able to transform matter even at the molecular level. There already exist nanotweezers capable of manipulating single atoms (IBM famously wrote their logo in single atoms) and this technology is inevitably gonna become better and cheaper.

(By the way, don't worry about environmental damage ruining the planet. Clean energy is growing exponentially. It will be so abundant that energy-generation will be a complete non-issue before the effects of global warming are really felt. The movement agitating to prevent ecological catastrophe is trying to apply 2010 technologies to solve 2050 problems, which is ridiculously stupid.)

At the rate things are going, we will be able to create an artificial replica of a human brain (artificial intelligence, in other words) by

  1. That is to say, by 2035 we will have both the computing power to replicate a human brain (10^16^ operations per second) and the scanning-technologies to see what is happening in there. The Blue Brian Project has already mapped around 5% of the computational networks of the brain. I don't know if this will lead to direct interfaces between consciousness and technology, where a semi-synthetic brain will give us the sort of superior data-retrieval and processing abilities of computers (2035 computers, not modern computers), but some people would rather merge with technology and become more intelligent themselves than have to hang out with robots that are superior in every way. I mean, think how smug those robots would be. The choice is between humanity improving its physical design or else accepting inferiority to machines. Personally, I don't even like the fact that my netbook can beat me at chess. Merging your brain with a computer ain't science-fiction; if the statistics are reliable, this will actually happen to anyone who plans on living another 25 years.\ And it is at about this point that singularity gets weirder. Much weirder. We've just seen that ordinary human intelligence can grow technology at a pretty absurd rate. But if superior intelligence is created, how much faster will technology accelerate? Maybe the speed of progress will approximate infinity (the time to accelerate technology will approximate zero) at this point around 2035. In other words, maybe everything will happen at once, up to the inherent limitation of reality for shit-that-can-happen.

Life-extension is a real possibility for any time in the next fifty years as well. It is harder to predict than something like computing speed, as we don't know what technologies will be used. But where we have medical nanorobotics, genetic engineering, increasingly accurate reverse-engineering of the human metabolism, it is hard to imagine all this having no effect on slowing down what we now call aging, especially when you consider that life expectancy has doubled in the past 100 years. Watch Aubrey de Grey drink beer and tell you all about his plans to stop aging, but bear in mind that Dr de Grey has not factored in the synergy between his field of biochemistry and other singularity-technologies like nanorobotics. It is definitely on the cards (as Dr de Grey has predicted) that the first people to live to be a 1000 years old are alive today, and may even be middle-aged today. It is not that we will completely cure aging in the next forty years or so; it is just that in the next forty years or so, we will find something good enough to partially reverse or slow down aging enough to buy another forty years, in which time we will see another forty years of exponential growth in technology, which allows us to see another forty years of progress - and so on until we have completely cracked the metabolic processes that lead to aging. If you're under forty now (maybe even if you're under sixty; who knows) there's every chance that your lifespan is effectively indefinite. Kinda puts your career plan into perspective, doesn't it?

Can we handle it?

The dangers of this technology are immense. If we could manufacture any DNA on our desktop, would people not use this from creating ugly viruses? If we could create nanorobotics, would some evil mind program them to eat human flesh? If we can manufacture anything at home, how long before people use it to make guns?

As for accidental hazards, our intelligence and technology improves, we will be able to handle (though not perfectly) the inevitable blunders. We may create accidentally create a killer virus, but if we have the technology to do that, we should also have the technology to cure it. Look at information technology: there have been both evil applications and bugs, but they get worked out and by-and-large the technology works for good.

The dangers are there, but I see two or three encouraging signs. One is that in the 65 years that we've had nuclear weapons, which are about as catastrophically dangerous as anything can possibly be, we have had enough sanity (just barely) not to do the stupid thing we might've done with them. The other thing is that as an abundance of know-how creates an abundance of water, food, education and the necessities of life, people will have much less reason to be pissed off; most evil is driven by acquisitiveness and is rendered obsolete is a situation of abundance. Another good sign is that, as people like Cristopher Langan and Stephen Wolfram apply computational theory to move us closer to an understanding of the true nature of reality, it will become harder and harder for people to deny our basic unity. But ultimately, there is only one solution. Technology holds the promise of infinite power, and there is nothing in the universe that can constrain infinite power - except for Love.

Can we use our technology to increase Love? Maybe humanity's fate ultimately rests on that single question. It is very heartening to say that the results so far from positive psychology, Buddhism and neuroscience suggest that the answer is Yes.

See also

Category:The Future