Surviving the Singularity

by Steve Alan Edwards

"Meat!...You were born meat and you will die meat. There is nothing I can do for you." --Diandra in Antibodies by David J. Skal. Just over the horizon are hurricane force winds of change-- "change comparable to the rise of human life on earth" according to mathematician and computer scientist Vernor Vinge, also the author of such science fiction classics as True Names and Marooned in Real Time. The cause of the change he envisions is "the imminent creation by technology of entities with greater than human intelligence" an event he expects to occur between the years 2005- 2030. With super human intelligence in charge of technological progress, progress itself becomes much more rapid and involves the creation of still more intelligent "beings" on an ever shortening time scale. "We humans," says Vinge," have the ability to internalize the world and conduct "what ifs" in our heads; we can solve many problems thousands of times faster than (Darwinian) natural selection. Now by creating the means to execute those simulations at much higher speeds, we are entering a regime as radically different from our human past as we humans are from the lower animals... From the human point of view this change will be a throwing away of all previous rules, perhaps in the blink of an eye, an exponential runaway beyond any hope of control." He calls this eyeblink "the Singularity" , a time when change occurs at such blinding speed that mere humans will be rendered obsolescent-- "the physical extinction of the human race is one possibility." But wait! Wouldn't it be really great if we, by packaging ourselves into a machine (or a machine into ourselves) could somehow achieve that greater-than -human intelligence, and become our own evolutionary successors? Wouldn't it be great if we could survive the Singularity? Meet the Transhumanists--an Internet -connected virtual community of futurists whose stated goal is self -transcendence through technology. International in scope, though few in number, transhumanists tend to be young, intelligent, and technologically literate - -often graduate students in neuro- or information science (see below). Along with Vinge, their intellectual heroes include roboticist Hans Moravec, artificial intelligence pioneer Marvin Minsky, nanotechnology guru K. Eric Drexler, and physicist/cosmologist Frank Tipler. Moravec and Minsky have argued for the theoretical feasibility of "mind-uploading" wherein a person's mind and personality could be emulated by a computer. Drexler has argued that the Singularity is even closer than we think, driven by --you guessed it- -nanotechnology, the science of creating objects by controlling matter on a molecular scale. Tippler's cosmological scheme holds that the universe is evolving into a giant supercomputer which he chooses to call the Omega Point- -but is, perhaps, indistinguishable from God. Included under the general heading of transhumanist is the Extropian Institute, a cult-like club of mostly Californians, whose cheerfully grandiose projects occasionally win them press attention. Extropians combine their transhumanist ambitions with libertarian politics and a philosophy of "dynamic optimism". Through the agency of e-mail, I assembled, a(n admittedly non-random) group for the purpose of determining the core beliefs, attitudes and activities of self -described transhumanists--three Americans and three Europeans. Two of the six are associated with the Extropians. Though they communicate through the 'Net, none of the six have actually met each other in, as one put it, "Meat-Space." What follows is an annotated and shamelessly edited panel discussion. A primary tenet of transhumanity is that something like the Singularity will actually occur. However, according to one panelist, the Singularity is "inherently polysemous",-- it means different things to different people-- (a new word for me, too), so my first question was: What is your definition of the Singularity? JKC> The singularity is a moment in time when things are changing so fast its impossible to predict what will happen next. ...The computational power we have with our puny brains made of jello and clumsy steam powered computers is pitifully tiny and is not increasing all that much. The computational capabilities we will have at a time near The Singularity will be huge and increasing at an astronomical rate, that's why we can't hope to have a realistic model about what life will be like then. EL> The Singularity, or The Grand Transition, is a developmental discontinuity, an ultimate event horizon beyond which predictability breaks down totally. EWF> The most general definition of the Singularity is the point in the future beyond which our ability to meaningfully plan our individual futures decays into noise. SM> I see the Singularity as a point in time beyond which progress of all types proceeds in ways we are incapable of predicting or of comprehending. Is the Singularity inevitable? JKC> Yes. EL> The Singularity is not at all inevitable. It can easily be prevented by a catastrophic global breakdown (Club of Rome scenario...) DGC> It's inevitable. Nothing short of a civilization-destroying catastrophe can delay it, much less stop it. Vinge himself sees the as Singularity as inevitable, given the feasibility of producing a greater-than-human intelligence. Asked if we could pull the plug, Vinge says, "No. There would be some researchers who would continue to pursue the goal, so the global answer is "no"." The human brain is a marvelous if imperfect piece of hardware, containing trillions of synapses--a massively parallel architecture more powerful than the largest extant supercomputer. But computers get faster, more powerful, and cheaper every year, whereas the brain's evolution continues at a glacial pace. Hans Moravec, in his book Mind Children, extrapolates that we will have a human equivalent supercomputer about the year 2010 and desk-top model costing about a $1000 by the year 2030. The growth of the Internet and object oriented software, however may nullify these projections. Some have argued that the 'Net is, in effect, a massively parallel supercomputer. "The Net is the computer," has become the corporate motto of Sun Microsystems. In that case, human equivalent hardware parity may be almost available now, if one could write an efficient program for an artificially intelligent "mind." Vinge foresees the possibility that "large computer networks (and their associated users) may "wake up" as a superhumanly intelligent entity." Alternatively, Vinge says, "computer/human interfaces may become so intimate that users may reasonably be considered superhumanly intelligent." Improving that interface is the goal of both the computer industry and neuroscientists. Voice activation and virtual reality are two approaches already available if not very efficient. Computers are being trained to recognize and respond to human emotional states (dogs can do it, after all). Brain waves have been used to control computer cursors and mouse buttons, and PC based EEG-like machines are already available for hobbyists. Peter Fromherz and Alfred Stett of the Max Planck Institute have succeeded in developing a silicon chip that can directly stimulate single neurons without damaging them and receive information back in the form of ionic nerve impulses. Can a true cyberpunk style brain-computer interface be far away? Whether or not the superintelligent catalyst has a human component, my transhuman respondents put the arrival of the Singularity within 10 to 60 years. Given the imminence of this wrenching, certainly chaotic change, what can society at large do to prepare itself? EWF> Propagate Buddhism. I'm not kidding. Buddhism is the only religion I know of that has evolved what I think is an approximately sane value system for the future, and it is the only religion I know of that asks its adherents to deal with the fact that change, personal change, is inevitable. JKC> Nothing. You and I may or may survive the Singularity, but society as we know it certainly will not. AS>Society must become much more flexible to survive even the beginning of the Singularity, since diversity, change and new problems will appear so fast that no current system could handle it. Just look at the confused reaction of society to the Internet and now scale it up a thousand times. EL> Duck and cover. Seriously, though, the majority (as usual) is not consciously aware of what is coming. SM> Barricade the doors and hide under the tables. Seriously, I don't think there is any point in speculating how society at large can prepare itself, because society at large will never accept the future existence of a Singularity. Vinge echoed the opinions of the transhumanists. "If the symptoms continue to become stronger, I think the idea will become a (much-doubted) commonplace among futurists. Even so, the actual event could come as a big surprise." Is the Singularity a good thing? SM> I certainly fear the Singularity, for I fear the loss of control that it may bring with it. But I cannot say it is a bad thing, for by definition, I cannot know what it will bring. All I can say is that is a scary thing. ...The technological expansion of my own intelligence may serve to keep me ahead of the game, such that I become the creator of the Singularity rather than it's victim. JKC> We are about to enter a period of gargantuan change happening in an astonishingly short amount of time and that is an inherently dangerous situation, it would be foolish to deny it. ...In spite of the dangers I admit I'm happy about the coming changes, we might survive it, and the alternative after all, is old age and death for all of us. Ah, the dream of immortality--to have your thoughts, dreams, and personality securely etched in clean shiny silicon, instead of its current implementation in nasty old carbon, with its propensity to rot. That is the basic lure behind the Moravec and Minsky idea of "mind uploading." How would it work? Proponents of uploading contend that the brain is a computer; therefore, in an extension of the Church/Turing hypothesis, the brain can be emulated by another computer. However, in order for an emulation to work, detailed information must be available about how nerve cells in the brain are interconnected, and whether synapses are stimulatory or inhibitory. Moravec presents a fanciful description in Mind Children of a robot surgeon whose nanometer sized fingers trace out human neural pathways in and then translate those into emulated neural networks, a layer at a time. The same robot surgeon appears in Harrison and Minsky's science fictional The Turing Option. Neural pathways can be traced now by heroic application of such tools as the Confocal Microscope (a light microscope capable of focusing at different planes in a relatively thick slice of tissue), or the Transmission Electron Microscope (which forms high resolution images based on the ability of samples to deflect electron beams). Already, the complete nervous system of a nematode has been detailed. (a nematode is a tiny worm-like organism of less than 1000 cells). Unfortunately, these techniques require that the brain be fixed, sectioned, and very dead before they can be applied. Joe Strout at the University of California, San Diego is one of the few neuroscientists (as opposed to AI proponents) who will admit to an interest in uploading. Already, he says, it is possible to "build a "black box" model of a particular neuron, based on its responses to any (injected) input with high precision." Strout is extending this technique by injecting dye into recorded neurons and then scanning their morphology into a computer. Eventually, he hopes to be able to be able to mathematically model nerve cell output from morphology alone, eliminating the need for stimulating and recording. Then, he says, "we would have a technique which could in principle scale up to any problem--even entire human brains." Inserting a little realism into the discussion, he estimates that "with current technology, scanning a single human brain would take 150 thousand years." Still he believes that within a few decades technology will be available that allows a reasonably complete description of human neural pathways. Vernor Vinge thinks that mind uploading is an attractive idea and will be feasible. He worries though about the "superhumanity issue. If you upload into something big enough, it's like a thimbleful of dye (your soul) lost in the Pacific Ocean." Consider the problem of time. If you've had a traumatic near death experience or taken enough recreational drugs, you probably realize that time sense is subjective; your perception of time depends on how fast your brain is working. No matter though whether you think that time now is clipping along at breakneck speed, or running like molasses--imagine if your brain worked 1000 times faster, the way computers do, even at current clockspeeds. The first year after uploading your consciousness into a computer would seem as an eternity compared to all your years as a human being. Your earlier identity would recede as though a fetal memory. My panelists were unanimous in their decision to accept uploading, despite the problems, although some expressed caution: >SM I am very fond of my physical self and want to make it last as long as possible. But I am aware that even if longevity research means my current body will last for ever without aging, it will undoubtedly 'expire' within a few hundred years from general wear and tear, and, possibly, mutation caused by background radiation. And, of course, there is always the possibility of an accident wiping me out. So I'd like several backup consciousnesses, which remain a second or two behind my current in-use self so that, when the time comes for me to be evicted from my body, I can call on a backup me to ensure my continued existence. JKC >Yes. If the scan of my brain was non-destructive then I would volunteer to be the first one to be uploaded, and would do so in a heartbeat. If as is more likely, the scan was destructive, then I would be worried about bugs in the technology that had not been ironed out yet and I might prefer that somebody else be the first test pilot. It wouldn't be wise to wait for long however, the first uploads would soon leave the meat machines in the dust. An alternative to "mind uploading" is the idea of hardware implantation, jacking digital components into your brain as popularized by William Gibson in Neuromancer. Already, deaf people have had their hearing at least partially restored by cochlear implants that transmit electrical impulses directly to the optic nerve. Similarly, "vision chips" are being tested at MIT on animals, and there has been at least one minimally successful human test. Microelectronics has been used to add motor control to prosthetic devices and, in some cases, sensory output from those devices. These medical attempts to repair the body might be the opening wedge in an expanding industry dedicated to extending the mental and physical functions of human beings through electronics. The body modification movement, which so far is limited to tattoos, steroid consumption, and threading rings and bars through every conceivable appendage, could get a tremendous boost from implanting math-coprocessors, hard drives, miniature video-cameras, infrared vision, cell-phones--the mind boggles at the possibilities. If Neuromancer type intelligence enhancements were available, would you take advantage of them, assuming you were assured of their safety? AS> Definitely. I already feel an intense need for coprocessors to handle parts of my short- and long-term memory, keeping track of various projects, visualizing and handling advanced math and similar stuff. JKC> Certainly! I can't think of any reason not to. Already, there have been low mumblings among the digital community about "neuro -hackers", underground practitioners of the digital/medical arts. Joe Strout, for one, is skeptical. "You can confidently ignore any such rumors. Progress is slow enough with the multi-million dollar equipment and the experienced researchers at labs such as mine and others; kids working in their basement are not going to accomplish much." In any case, most transhumanists see this "cyborgization" of human beings as a kind of stop-gap measure, allowing them to keep up temporarily with the Singularity. "This would be only the first step ... I would choose this option if true uploads were delayed or unavailable (EL)." The Singularity will be, perhaps, the ultimate culmination of human cultural evolution. In chronological time, most of that evolutionary history, perhaps a million years, was involved in simple tool construction, and in perfecting language as a means of communication. Language allowed also for the codification of thought, a way for a person to communicate with his own mind. Technology accelerated a little faster when language became written, and speeded up quite a bit more when Gutenberg mechanized the process. Textbooks were written and education on a mass scale became possible. Most of the scientists, engineers, and technicians who were ever alive are still alive, with all their collective brainpower. Computers have extended their computational power, and of course, the Internet has allowed cheap and rapid communication between scientists as well as intercomputer discourse. How much machine power is required to replace the scientists entirely? Transhumanist JKC quotes K. Eric Drexler's calculation that "10^38 machine instructions" would "simulate ALL the brains (non-humans included) that ever existed ...since brains were invented in the Cambrian Explosion 570 million years ago." The simulation would require a "Nanocomputer the size of a large present day factory and using no more power" about two years to complete. It is cosmic considerations like these that drive transhumanists to the conclusion that the further evolution of man involves something that is, well--post-human. The ultimate goal of a transhumanist, by definition, is to attain post-human status. One of the few points of disagreement among my panelists is whether elevation to this happy condition will be available to all. Do you think that a cross-section of humanity will make the transition to post -human or only the technological elite? AS> Unfortunately, it is likely that many people will not become posthuman. Some will not want it, or dislike it due to religion, coservatism or fear. But many people are simply not aware of the possibility, not aware of their own ability to grow and develop. EL> If there is a single altruistic post-human individual (Does he/she/it have the Buddha nature?), he can post-humanize the rest of us with the equivalent of a snap of the fingers. EWF> I think that most but not all people younger than 30 years old today will end up becoming something more than merely human. SM> Probably only an elite, but it won't be the technological elite, it will be the financial elite. The technology required to reach posthumanity will hold such a premium that it will be available only to those who develop the technology, those they choose to give it (family, friends etc), and those that can afford it. Time is getting short. Most of the Transhumanists expect the Singularity to occur in their lifetimes. Neural implants are not yet available and mind -uploading is still in the realm of science fiction. My final question to my panelists, therefore, was: What are you, personally, doing that will allow you to make the transition from transhuman to post-human? JKC> Trying to stay alive until the Singularity and planning to make use of cryonics in case I cannot. (Cryonics is the science of ultra-low temperatures. In this case, it refers to the practice of freezing people at death, in the hopes of resurrecting them after medical science advances). EL> Trying to choose constructive (Singularity-advancing) topics for my research. Engaging in installation of basic Kraut (German) cryonics infrastructure. Acquiring sufficient financial means to broaden my future light cone. EWF> I've been studying ethics and epistemology for many years, because I feel that anyone arrogant enough to want to attain the powers that we suspect will be characteristic of posthumans had better make sure they've asked and studied all the questions they can about their ethics and motivations. ...Basically, what I'm doing right now is preparing my mind for where I think the center of the action is going to be ten years in the future. SM> Keeping abreast of the technologies. Starting a Masters degree in Neuroscience. Engaging in discussions about posthuman ethics and society. Smiling nicely and making friends.