From: Ralph Cerchione (figment@boone.net)
Date: Sun Jul 17 2005 - 23:00:50 MDT
And here's one further article, from:
http://futureimperative.blogspot.com/2005/05/host-of-variables-part-i-computers.html
Ralph
---- Recently I chimed in while others were debating at exactly what point computers will be sophisticated enough to simulate an entire human body at an incredibly precise scale... and then run medical trials on their virtual human. I commented... "I haven't looked too closely at the numbers for a credible, molecular-level simulation of the human body -- at least, not in quite a while. But I'm willing to accept that nano-scale modelling of the human body may be possible in the not-too-distant future, even if the task turns out to be extraordinarily complex. "Why? Because no matter the pace of computer hardware development post-silicon in 2020 or 2030, it's quite possible that by 2010 someone will say 'Hey, let's hook together a million Playstation 5s, and see what kind of processing we can do with that setup!' "In other words, the outer limits of what supercomputers are capable of are not completely limited by Moore's Law, but by the willingness of innovators and large organizations to put together massive yet efficient computer architectures. Such as the massively parallel project I just suggested. "And if one million Playstation 5s doesn't sound that impressive, imagine linking up five million, or fifty. Many massively parallel architectures are marked by the fact that no one has spent billions to construct them, and by, say, 2015, both our resources and our programming savvy should be much greater. Making ridiculously powerful designs possible for even smaller governments and less powerful corporations. And thus making possible implausibly advanced research in biotech and other key fields. "And the above option represents only one potential computational innovation. Quantum computers, DNA processing, alternative substrates and basic nanotech all offer intriguing options -- and at least two of them are revolutionary in their potential." And all of these variables represent only a few of the factors influencing just the field of computational hardware design. There are many more major variables that could impact the research and development process in this particular field. For example, what if the excess talented engineers in China and India get involved in this research in a far more substantial way, thus broadening the research base and increasing the number of top notch innovators working on key problems (as Microsoft started doing years ago in software when it located its third global research center in China). Alternatively, what if there is a war between China and Taiwan? That would critically disrupt the international computer fabrication supply chain running down the side of East Asia and Southeast Asia. The chain would survive, by moving to other countries, but major players in the computer industry could be displaced from their positions, with unforeseen implications for the tech world, and tech researchers. Or there could be a huge government push (by the U.S., the EU, Japan, China, India and/or others) to develop nanotech, AI or simply supercomputers optimized for the most cutting-edge biotech work. Whether such a push would be wildly successful or utterly disastrous is beside the point. Whatever happened, enormous financial, human and technical resources would be diverted into a particular plan for achieving the desired breakthroughs. And these are only political scenarios. But remember whenever someone starts telling you about the steady progress of Moore's Law over the decades... it's not a law of nature. And if political decisions can have a huge influence on near-term technological development, the interactions of seemingly unrelated fields could have an even greater impact. Imagine what would happen if you had a computer efficiently sifting through medical journals and databases looking for the most ideal drugs, and developed an incredibly powerful nootropic (intelligence-boosting) drug... which was then taken by biotech and other researchers. Or if a new system of rudimentary nano-fabrication made supercomputers many generations smaller and more powerful, and enhanced that drug-sifting software, and also put immense processing power in the hands of each of those researchers, at a trivial price. Or if massively networked/parallel supercomputers using millions of processor nodes kicked both gene therapy and germline engineering into high gear and made genetic superhumans possible... and cheap enough to convert whole populations as desired. Or if several advanced and developing countries used basic mindtech (such as floatation tanks and biofeedback) and accelerated learning and creativity enhancement methods to train a few hundred million genius to slightly better-than-genius scientists and engineers... and a couple billion immensely skilled technicians, entrepreneurs, creative knowledge workers and other "willing (and talented) hands" to exploit the fruits of their genius. Futurists often debate what phenomena could trigger that wave of accelerating change called the Singularity -- a point at which change -- particularly technological change -- is occuring at such a rate that we can no longer speak with any degree of certainty about what the future may hold. Any of the above items could potentially move us into a Singularity by way of what some would call a "soft takeoff." That's the version where things are changing at a pretty profound level, but your local state, province or country doesn't get rendered down into molecular-scale processors by rogue nanoswarms in the next couple of hours, or probably even the next couple of months or years. Nor do you suddenly "wake up" inside a virtual utopia, your body having been reduced to solid state electronics while your consciousness was instantly "uploaded" into a paradise in the Matrix -- because it was a far more efficient use of matter to reduce you to software when your body was taken. The "hard takeoff" is the one in which that could happen, but it's probably far too prosaic and low-key to be credible. And whatever really happens will probably happen sometime in the next ten hours to ten nano-seconds. It was nice knowing you. Goodbye. What's my point? Well, among other things, there's an argument to be made that a soft takeoff that leans upon biotech and "cognotech" may be easier to manage (not to mention survive) than a hard takeoff dominated by superintelligent AI of unknown and possibly alien motivations. What's more, whether or not we get an official Singularity, there are several different fields of research which could help humanity develop "more-than-human" skills and intelligence, thereby transforming civilization forever. What's more, we're far more likely to make huge strides by pushing several major fields simultaneously, especially if those fields are capable of nudging us forward even with merely modest, incremental successes. A nootropic drug here, a genetic analysis of intelligence there, a nationwide retooling of schools... everywhere, and before you know it, all kinds of factors enhancing our research base and our ability to exploit new technologies are coming into play. Thereby accelerating the rate of accelerating change. In the second half of this article, I will be addressing some basic initiatives that could trigger a "soft takeoff" Singularity within about 10 to 15 years. There are more than you might expect. http://futureimperative.blogspot.com/
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:51 MDT