Future Shock Level Analysis
Michael Anissimov :: Feb. 2002, updated Jan. 2004


The following is an expansion on transhumanist Eliezer Yudkowsky's "Future Shock Levels" idea, a rough model that seeks to generalize varying levels of "future shock" and concepts or technologies characteristically associated with these distinct levels.  The concept of space travel, for example, possesses more "future shock value" than the concept of air travel, although both are concepts now familiar to the bulk of 21st century humanity. An example of a statement likely to incur future shock in the contemporary layperson might be: "barring any major disasters, there's a strong possibility that medical nanotechnology will eventually be used to cure all diseases and extend human lifespan indefinitely". This may be true, it might not be, but if it shocks you, then you aren't ready to think about the possibility logically. Future shock is a genuine problem in our world, causing millions to withold support from humanitarian projects that hinge on a thorough understanding of technological futurism. The problem of future shock continues to get increasingly serious as technological progress accelerates; human beings can only absorb so much future shock so fast. This essay explores the phenomenon of future shock and some possible underlying causes, then suggests potential solutions and presents a more detailed version of Yudkowsky's original scale.


Future Shock

(Although the term "future shock" was originally coined by Alvin Toffler in his book of the same name in 1970, this essay primarily focuses on Yudkowsky's extension of that idea into distinct "Future Shock Levels".)

Future shock is defined by Yudkowsky (1999) as being "impressed, frightened, or blindly enthusiastic" towards the future, or specific social or technological advances. On the whole, it that those who are more familiar with and sentimental towards the technological and social states of the past (usually older people or those who live in rural areas) are, on average, more highly predisposed to being irrationally frightened than blindly enthusiastic of the future. Conversely, we should expect those more familiar with and sentimental towards the technological and social states of the present or future (usually younger people or those who live in urban areas) are probably more at risk for blind enthusiasm than irrational fear. From a cursory review of historical responses to new technologies or social arrangements, it seems that on average, the "irrational fright" response is significantly more common than the "blind enthusiasm" response.

It is commonly recognized that the primary distinguishing feature separating the humans of today from the humans of 50,000 years ago is the sophistication of our technology. The past 50,000 years have only represented miniscule steps in mankind's biological evolution. Evolution works on extremely long timescales. It requires strong selection pressures and tight breeding circles for new, advantageous mutations to arise and spread. When evolutionary selection pressures were globally dampened by the forces of reciprocal altruism, better nutrition, shelter, and medicine, biological evolution massively slowed (from its already incredibly slow pace!) Factor two: when human breeding circles expanded from an average size of 100 prospective mates to 1,000 prospective mates, advantageous mutations began to have more trouble becoming species- or community-typical, as they tended to be diluted in the ambient static of quantitative variation among traits that is inherent to larger communities of organisms.

Biologically, humanity has mostly stayed the same. But ideologically, memetically, we have progressed. And what is largely responsible for that? Many would agree that it is our technology. Humanity's ancestral niche was the African savanna, where we lived in tribes of approximately 200 individuals, engaged in hunting and gathering exercises year in, year out. The root cause of the global migration of Homo sapiens, our ability to artificially adapt to and manipulate the environment, and the formulation of cooperative groups with significantly more than 200 members was the invention and development of technology. Without technology, humans would have no particular impetus to develop substantially varied types of culture, or to formalize the pursuit of knowledge, or to engage in artistic, philosophical, scientific, or large-scale humanitarian activities as we know them today. We might have continued on as unusually intellectual chimpanzees for thousands or even millions of years, reproducing and dying and suffering without tangible progress of any kind. The bottom line: evolution is basically over, only ideas and technology change. Ideas mostly change as a result of technology, although there are subtle feedback effects at work.

New ideas inevitably emerge with new technologies. Studies have shown that literate individuals combine and analyze concepts in qualitatively different ways than illiterate people do (Klawans, 2000). Without the invention of writing, these concrete neurological differences never would have come into being. A printing press that allows a middle-class person to directly spread their ideas to hundreds or thousands of people gives rise to deeply different memetic dynamics than a society based on word of mouth alone. The possibility of worldwide travel through cheap aircraft shattered cultural barriers like never before. The invention of the atomic bomb changed the face of war forever. Further examples number in the millions or billions. As technology continues to radically change our world, its singular importance remains undeniable. Although we tend to delegate most of our attention to the specific attitudes, behaviors, and thoughts of other humans and depreciate the importance of their inanimate contexts, it should be remembered that we humans are operating on a richly structured, largely imbalanced technological playing field worthy of explicit attention. Oftentimes, the analysis of the underlying technologies enabling human agents is more pragmatically valuable than specific attention towards the qualities of the agents themselves.

Humans have a proven history of failure when it comes to predicting future events, especially events which will influence them personally (Slovic, 2002). We are much better at predicting trends, which are grounded in past data and may hold reliably for decades or even centuries. Some trends necessarily culminate in events. For example, one pervasive trend throughout human history has been an increased capacity for information exchange between any two individuals. Prior to the invention of the Internet, this trend was clear. With telephone, fax, and snail mail operating as the best communication methods of the time, an eventual advance of some sort was highly probable, and this advance did indeed occur, and was predicted correctly by thousands of futurists, science fiction writers, technologists, scientists, and others. Today, it seems astronomically likely that the maximum possible information flow between any two individuals will continue to increase (barring any huge disasters), through the mediums of larger servers, ubiquitous broadband, peer-to-peer networks, free long-distance, and so on. If no disastrous events intervene, the longer-term future will allow the actual transfer of cognitive content through advanced nano/neurotechnological communications media.

Some famous quotes of failed predictions:

 

"Computers in the future may weigh no more than 1.5 tons."
- Popular Mechanics, 1949

"I think there is a world market for maybe five computers."
- Thomas Watson, chairman of IBM, 1943

"But what... is it good for?"
- Engineer at the Advanced Computing Systems Division of IBM, 1968, commenting on the microchip

"There is no reason anyone would want a computer in their home."
- Ken Olson, president, chairman and founder of Digital Equipment Corp, 1977

"This 'telephone' has too many shortcomings to be seriously considered as a means of communication. The device is inherently of no value to us."
- Western Union internal memo, 1876

"The wireless music box has no imaginable commercial value. Who would pay for a message sent to nobody in particular?"
- David Sarnoff's associates in response to his urgings for investment in the radio in the 1920s

"The bomb will never go off. I speak as an expert in explosives."
- Admiral William Leahy, US Atomic Bomb Project

"Heavier-than-air flying machines are impossible."
- Lord Kelvin, president, Royal Society, 1895

"Airplanes are interesting toys but of no military value."
- Marechal Ferdinand Foch, Professor of Strategy, Ecole Superieure de Guerre

"Man will never reach the moon regardless of all future scientific advances."
- Dr. Lee De Forest, inventor of the vacuum tube and father of television

"Everything that can be invented has been invented."
- Charles H. Duell, Commissioner, U.S. Office of Patents, 1899

Why are humans so poor at predicting future events, and the invention of future technologies in particular? Part of the cause goes back to how humans were originally designed by evolution. Our minds were constructed to solve certain adaptive challenges, and any mental abilities emerging outside of specific solutions to these adaptive challenges are probably general-purpose mechanisms that just happen to be applicable across a broader range of domains than usual (Cosmides et al, 2001). As a result, humans have few innate mental tools for analyzing the plausibility or impact events more than a few months distant, especially phenomena as complex as the arrival of futuristic technologies. We are not specifically adapted to realistically assessing the likelihood of a particular technology working successfully or unsuccessfully, or the social impact of the distribution of that technology. In the ancestral environment, the environment that humans originally adapted to, new technologies were invented rarely if at all. The main factor that dictated the fates and fortunes of human beings would be their short- and medium-term interactions with others. Although human beings do display a certain degree of logical flexibility, complete transformation from default human reasoning to a theoretically optimal decision strategy is not to be expected (Koehler, 2002). Humans have several species-wide innate mental weaknesses, and a handicap in the area of accurate technological forecasting is one of them.


Calibration Strategies

Debiasing the negative effects of future shock can be attempted in several ways. One requires becoming familiar with a wider range of the potential possibilities the future might offer. Hollywood and television futurism are only of limited use in this respect; these forms of mass media are customized for the elicitation of ratings, not scientific accuracy or futurological plausibility. Higher quality sources would include books and magazines associated with currents of thought whose past predictions have been confirmed. For example, less than a decade before the creation of Deep Blue (a chess-playing supercomputer), chess experts commonly claimed that no computer would make it past a certain level of expertise in chess. AI experts asserted that these chess experts were almost certainly wrong. When Deep Blue proved to be an intense challenge to world chess champion Boris Kasparov, the AI experts' predictions were confirmed. This is just an isolated example - currents of thought must build up multiple trials of confirmed predictions before assigning them special validity is warranted. Conversely, currents of thought which have displayed multiple failures in prediction should be negatively weighted. This would strongly depreciate predictions for certain future events; for example, the emergence of psychic powers or the arrival of extraterrestrials.

Many commentators on the possibility of man-made flying machines prior to the Wright Brothers were struck by future shock. They observed the performance of past technology, the mythological history of human flight, and the heavy weight of proposed airplanes in comparison to birds, and concluded that would-be aeronauts were simply wasting their time with a physically impossible endeavor. Interestingly, the physical theory required to appreciate the plausibility of artificial flight was widely available at the time. Newton's laws of physics can be used to show why objects heavier than birds are indeed theoretically capable of flight, and by 2004 been used for exactly this purpose many thousands of times. This could have been mathematically confirmed by skeptics prior to a successful flight, but it wasn't. As a result, widespread skepticism of the possibility of flight persisted until these beliefs were experimentally disconfirmed. It shouldn't have needed to so that long. "Learning from our mistakes" in the domain of technological futurism means paying close attention to theory and closing ourselves off from extraneous factors such as the surprising-looking comparative weight between proposed airplanes and birds.

The neglect and distrust of scientific models in futurism is not so shocking considering the neglect of science in general among the public, even among intellectuals. Direct, personal-level interactions between people are inherently fascinating to read about, think about, and discuss - the kind of knowledge that would have helped our ancestors survive on the African savannas. (Struggles for power within hierarchies, the violation or preservation of social taboos, displays of personal resources, the phenomenon of romance, etc.) Science requires explicitly disengaging from these age-old preoccupations and analyzing phenomonena on more abstract levels. Try turning on the TV or reading a popular magazine. Soap operas are a distilled-for-TV version of romance, and the evening news is always heavily politicized. Romance and politics are much more emotionally engaging than science, technology, futurism, and mathematics, and require a more modest level of intelligence to understand and participate within. True scientific futurism is less about exhuastively understanding traditionally fascinating interactions directly between humans, and more about understanding the relationship between humans and the technology we create, including the way that technology modulates everyday human interactions. However, futurism in popular society is merely a tool to entertain, and popular society's overall beliefs end up regulating which books, television programs, messages, and people receive attention. Typical types of human interaction are as old as our species; they’re largely constant relative to technological changes. This asymmetric focus on traditional human activity discourages accurate forecasting of our future. (Perfectly accurate forecasting of the future is clearly impossible, but if your accuracy rate jumps from, say, 30% to 31%, that small change could translate into thousands of lives or billions of dollars saved.) The point: calculated rationality and scientific sobriety can be expected to produce predictions of the future that are actually better than the guesses of casual laypeople. Not perfect, but consistently better.

Our human intuitions about societal change come from models that judge how present-day human behavior might or might not lead to a future state of human behaviors. In a technological civilization, the changes are not primarily directed by human behavior changing human behavior, except in the sense of human engineers creating new technology, which leads to unintuitive future states. Millions of years of evolution is telling us this is how it works:

When it actually works more like this;

Discarding the former model for the latter is far more complicated than simply recognizing it as better. The vast majority of human thinking is automatic; changing your underlying thinking processes about the future requires consciously recognizing when you are thinking in human-behavior-causing-human-behavior terms, and this is not easy. Technologies that work reliably - cars, telephones, houses, and computers - quickly become transparent, integrated into our usual daily routine. We just forget they're there because they're relatively seamless. However, this can lead to poor forecasting when we fail to account for the entire humans + technology meta-system, instead focusing on humans, human passions, and the salient technologies that just happen to arouse human passions.

Take the technology of cloning for example. Joe Six-Pack is given the task of predicting a window of confidence for the technological feasibility and arrival time for human cloning, in addition to giving a short description of the sociological effects of the cloning process. Sadly for Joe, the majority of fictional, popular literature on cloning has involved the immediate creation of adult copies of a given human being, often gifted with the precise set of memories the human possesses at the instant of “cloning”. This type of “cloning”, while technologically extremely difficult (requiring advanced nanotechnology and molecular scanning procedures rather than simply biotech), constitutes a “literary attractor” – it’s fun to write and read stories that conform to it. They sell and people like them. The reason is that the audience doesn’t know enough science to be sufficiently annoyed to compel large portions of the population to turn to more accurate sources of fiction. Instant-copy cloning just seems, intuitively, what cloning should be like. Yet the real technology of cloning will surely not correspond to Joe's model, and generalizing from scientifically unrealistic fiction is one of the most distorting forecasting pathologies of all (Yudkowsky, 2003). Joe's attempts at a remotely realistic estimate are doomed to fail. We can test the quality of different futurological "currents of thought" by observing how well their predictions stand up to the way events in the real world are actually unfolding.

The continuous creation, consumption, and sharing of human media creates a sort of memetic ecology – an arena where replication and artificial selection takes place on a grand scale. Imagine a story with ray guns and matter duplicators, yet a blatant lack of any type of flying machine – be it plane or spaceship. People would be confused. “How can this civilization have ray guns and matter duplicators but lack even the simplest propeller plane?” they might say. Such a story would flounder in the competitive environment of media creation and distribution, giving way to a more exciting story with ray guns and spaceships of all sorts included. This artificial selection process will vary in intensity based on regularities in the annoyance levels and specific interests of the typical consumer, dictating the shape of clusters in the scientific accuracy of popular fiction. The most powerful literary attractors currently lie somewhere in the space between cautious scientific realism and outright nonsense. (Example: Michael Crichton's "Prey".) These stories often contain odd mixes of popular, scientifically plausible technologies with popular, scientifically implausible technologies.

“It’s just a movie!” many people might say to visibly unrealistic science fiction. But not so to science fiction so profoundly absurd that it rudely violates the common sense of the audience, loses sales, and quickly passes out of existence entirely. In fact, most of this type of science fiction is never created to begin with, because whoever is writing the scripts probably has a level of technical knowledge about equivalent to the audience, and can empathize with what they expect and would or would not be annoyed by. This “boundary of annoyance” creeps forward on a daily basis, as the aggregate population gains better common sense in the areas of science and technology, what we know about human nature, and how they interact with one another. The average scientific accuracy of our “science” fiction slowly creeps upward, but often remains decades behind what has been obvious to scientifically literate technologists, futurists and researchers for many years.

The idea of cloning as creating fully-made adult copies of the clonee has not yet passed into the level of complete and widespread boredom and annoyance. Real cloning is quite complicated, and requires a knowledge of biology beyond what most of us learn in high school. Since most people do not continue to read about biology after graduating high school, they never hit the threshold of annoyance that reacts to this depiction of cloning. In many areas of the world where education is even poorer, and religion is ubiquitous, we can hardly expect the majority of the population to understand even the basic principles of genetics and evolution. If asked about the likelihood or potential impact of cloning; they have absolutely nothing to work with. There is no traction to create a prediction of any sort, and there can never be until the underlying science – biology, physics, computer science – are studied in sufficient detail. If the science is absent, traction will eventually be created based on rumors, first impressions, and the degree to which the proposed technology is representative of past experiences or thoughts (Citation).

Summary;

Begin forecasting by spending more time carefully examining the various proposed possibilities.
If any of them shock you, you are not prepared to think logically, and must desensitize yourself first.
Irrational fear of future advances is relatively more likely than blind enthusiasm.
The main difference between the past and the present, present and future, is the technology.
Human nature is relatively constant, but changes slightly based on the ambient level of technology.
Technology can strongly change human behavior, but humans do not deeply change technology.
Reliable underlying technology can seem transparent, but it is always important to take into account.
Be scientific. Read lots of science. Be scientific. Read lots of science. Be scientific. Read lots of science.
Avoid making generalizations from fiction or science fiction at all costs (Yudkowsky, 2003a).
The following points are so obvious that it's almost embarassing to mention them:
Don't fear the future just because it's likely to be different.
Don't underestimate the probability of change because change intimidates you personally.
Don't love the future just because it's likely to be different.
Don't overestimate the probability of change just because you want it.

Disclaimer: Viewing the following future shock levels (FSLs) as opposing social factions or exclusive clubs is of course tribalistic, and strongly discouraged. The FSLs are not a justificiation for discrimination or elitism. Just because the really low shock levels sound pretty awful doesn't mean that the overall scale progresses from "better" to "worse". You're supposed to make that judgement for yourself. Like religious and cultural associations, many people will probably feel most comfortable with the FSL they were raised or educated at.

Now that we've got all that futurism out of the way, here are the Shock Levels!

Future Shock Levels

SL-2 (negative two):
Major portions of developing countries.  This SL represents familarity and comfort with the technology of a century or two ago, and not much more.  Due to a lack of communication technology, information sources are limited to a few hundred people, likely possessing similar worldviews.  These conditions can encourage the emergence of false or destructive ideas. Daily life is often accompanied by a large amount of difficult, repetitive work and little return for the work done. SL-2s may avoid medical assistance or unfamiliar foods even if their lives are at risk. The vast majority SL-2s will be illiterate. I estimate 500 million or so SL-2s worldwide, with tens of millions thankfully moving to SL-1 or SL0 per year.

SL-1 (negative one):
A great number of people who live in rural areas or developing countries.  Barely exposed to modern technology, WWII-era technology is likely to be prevalent. Automobiles, engines, radios, and so on may be the forefront of available technology. Since some of these technologies must be purchased in areas heavily populated by those at higher future shock levels, some exposure is inevitable, and many SL-1s worldwide are rapidly becoming SL0. (Especially youth.) These individuals would probably have difficulty imagining what an "Internet" would be like without an elaborate and extended explanation. SL-1s are distinct from SL0s in that they would be "shocked" by SL1 and downright confused or scared of SL2 and above.

Sadly, due to communication technology limitations and poor education, these folks will probably have a fundamentalist bent, either religiously or ideologically.  There is a good chance they will be apprehensive towards or entirely unaware of modern day technology such as computers and other modern-era tools. Since all the technology they are familiar with is relatively crude, simple, and somewhat dehumanizing, the prospect of using better technology to improve their standard of living may sound somewhat implausible. Many concepts are in easy-to-understand black and whites, no continuities or "fuzzy thinking".  Natural and artificial are entirely different and seperate.  So are "science" and religion.  God and Man.  And so on. A good chunk of SL-1s will be illiterate and few will progress past the equivalent of a high school education. This isn't their fault, and it doesn't make them bad people. But they are suffering. SL-1s, I am guessing, make up the majority of humanity; maybe somewhere around 3.5 billion people. SL-1s have steadily been adopting SL0 ideologies and lifestyles en masse since WWII, with very few reverting to SL-2. As populous countries such as India and China are rapidly industrializing, hundreds of millions of new people per year are getting access to modern technologies.

SL0:
The “legendary normal person”, as Yudkowsky puts it.  Original description reads, "The legendary average person is comfortable with modern technology - not so much the frontiers of modern technology, but the technology used in everyday life. Most people, TV anchors, journalists, politicians." These people may have access to the Internet and assorted gadgets around the house, such as a modern TV set, digital clock, automatic shaver, etc.  Quite often possesses the hard dichotomous view of technology and humanity which makes it drastically difficult to step up to Shock Level 1 or Shock Level 2 ideas.  Daily conversation may indeed bring up technology a fair portion of the time - bearing in mind technology has already saturated most of the developed world. Technological improvements will generally be viewed as incremental improvements in efficiency or entertainment, but rarely as revolutionary or world-changing. An ambient distrust of technology may be accompanied by a notion that the broad outlines of technological progress are manipulable to human agencies (including the ability to "pull the plug"), when in fact they are becoming rapidly autonomous and human-independent.

These individuals generally assume that the lives of the people in each successive generation will be essentially the same as their previous generation, with a slight improvement in science and technology and a minor increase in lifespan and quality of life.  Untraditional technological advances may be approached with distrust or suspicion, often manifesting itself in the "Precautionary Principle", which demands that researchers convince major portions of the population to explicitly advocate the development of their chosen technology. As the pace of progress continues to visibly accelerate, many will polarize to technophilic and technophobic ends of the belief spectrum.  But for now, they primarily believe what they see on TV or newspapers (an odd mix of technophilic and technophobic memes apparently distinct to our time period and culture).  Seems to be the default level for most of the developing world.  Jetsons-esque visions of the future. Maybe 2 billion folks or so, with millions per year migrating to SL1, and very few moving to SL-1. SL0 seems like a strong, large, stable ideological attractor for most present-day humans within technological societies.

SL1:
A typical forward-thinking or well-educated person.  The original scale describes SL1 as "Virtual reality, living to be a hundred, "The Road Ahead", "To Renew America", "Future Shock", the frontiers of modern technology as seen by Wired magazine. Scientists, novelty-seekers, early-adopters, programmers, technophiles." SL1s are those who stay relatively current with technological and scientific advances. As stated in the original description, this level overlaps strongly with the portion of society containing researchers, programmers, tech investors, internet roleplayers, technology enthusiasts, and other science-and-technology-minded people.  It seems that if a "normal" person is exposed to enough technology and futurism in the right way, they will eventually become SL1s. Relatively conservative extrapolations of future progress in computing, robotics, the Internet, and biotechnology are represented here.  Extension of average lifespan to 100 or so, robot maids, and the like.  May entertain ideas of aliens, interplanetary colonies, antimatter rockets, and so on (not in an exclusively fictional sense). Every time a space mission or surprising technological advance is widely covered on the news, a few ten or hundred thousand SL1s will probably pop into existence.

SL1s may be able to point out specific physically unrealistic or scientifically implausible aspects of science fiction stories. SL1s prefer speculation and discussion of the future of mankind to thoughts of a religious afterlife. Anthropomorphic models of future artificial intelligences and transhumans will be inevitable, if such models exist at all. The "linear intuitive" vision of the future is more likely to be held than the historically realistic exponential view. SL1s will probably be relatively bibliophilic and intellectual, and most likely young or middle aged (but not always!) SL1s are probably not incredibly rare - we can estimate around 10 million or more worldwide. (Roughly one six-hundreth of the world population, or about one two-hundreth the population of developing countries.) Tens or even hundreds of thousands of SL1ers will reach SL2 per year, many of them youth who grew up with SL1 parents, peers, media, or teachers. Fans of mass media sci-fi such as Star Trek and Star Wars will often be SL1s.

SL2:
The average science fiction fan.  Yudkowsky's original description reads "Medical immortality, interplanetary exploration, major genetic engineering, and new ("alien") cultures." A common pasttime of this group might be speculation and discussion of extraterrestrial civilizations and advanced technology such as superlasers, starships, swarms of nanobots, human-equivalent (but rarely transhuman) AIs, and major bodily augmentation.  Realization of additional implications of biotechnology (clinical immortality and genetic engineering), computing (ubiquitous computing, 3D displays), robotics (elimination of manual labor) are inevitable.  The idea of extreme plentitude and the potential elimination of poverty and illness might come into play.  Starships, antimatter, cloaking; like many other future shock levels, this one features an unusual mix of physically realistic and unrealistic technologies.  Many people who are casually interested in nanotechnology would probably fall into SL2, as would many life extensionists. SL2s will have solid ideas about scientific feasibility and be aware of how modern-day technology works on a level significantly beyond the layperson. Scientists and engineers.

As their futurological models develop, SL2s may quickly find themselves unsatisfied with mainstream futurism and science fiction.  Major technological advancements don't seem as far away to this person than to most, possibly due to an adoption of the historically realistic exponential model of progress as opposed to an intuitive linear model.  Anthropomorphism, the hard distinctions between artificial and natural, matter and software, virtual and real, human and non-human, begin to fade. Physicist Michio Kaku would be a typical example of futurist thinking in line with the SL2 platform. SL2s may hold transhumanist philosophies which are generally biotech-centric. Perhaps one million or more individuals, each potentially receptive to SL3 or SL4 memes. Somewhere between a thousand and fifty thousand SL2s move to SL3 every year.

SL3:
Serious transhumanists, mostly.  Original description reads, "nanotechnology, human-equivalent AI, minor intelligence enhancement, uploading, total body revision, intergalactic exploration", although it seems appropriate to think of nanotechnology, human-equivalent AI, and intergalactic exploration as SL2 concepts. This is because they are actually quite common and mainstream nowadays, and familiarity with these concepts does not denote what is usually meant by "SL3". (According to the common usage I've observed, anyway.) SL3 represents a highly relevant and interesting threshold.  Transhumanists seek to improve their mental, emotional, and physical capacities by changing the physical structure of our bodies and brains. For better or for worse, SL3 contains the real mind blowing stuff.  Sophisticated nanotechnology, total body/mind revision through nanomedicine, uploading, potential posthumanity, Von Neumann probes, intergalactic travel, expansion throughout the universe, Dyson spheres, and the faint beginnings of nonanthropomorphic portrayals of AIs and other nonhuman intelligences are all common. SL3s tend to be scientifically rigorous and heavily bibliophilic.  Thoughts regarding SL3-level concepts may be prompted by readings in fiction or non-fiction, although the latter is significantly more common.

Most transhumanists want to live forever under the best possible conditions, expanding their bodies and minds, essentially becoming godlike by today's standards.  They want to engineer out negative emotions and enhance their own intelligence. SL3s probably realize that their brains are potentially hardware-extensible, and look forward to enhancing and expanding the physical substrate underlying their minds. SL3s often display an interest in memetics, advanced evolutionary thinking, and cutting-edge philosophy of science. Well-known science fiction authors discussing SL3 concepts would include Greg Egan, Vernor Vinge, Damien Broderick, and John Wright. Well-known non-fiction authors discussing SL3 concepts include Hans Moravec, Ray Kurzweil, Ed Regis, Hugo de Garis, and Bart Kosko. The popular Better Humans website features many interesting stories on SL3 and SL2 concepts. The boundaries between SL2 and SL3 can be fuzzy at times, but it seems safe to say that there are probably between 1,000 and 100,000 SL3s in existence, with between 100 and 1,000 moving back to SL2 every year, and between 5 and 20 migrating to SL4. (The backslide figures were revised downward by an order of magnitude after additional input from the SL3 community.)

SL4:
Singularity advocates and analysts. Sysop Scenario, singletons, deep thought about the implications of uploading and virtual realities, Jupiter Brains, Apotheosis, Alpha-point computing, evaporation of the human ontology, complete mental revision, recursively self-improving artificial intelligence, that sort of stuff. Many SL4s are interested in artificial general intelligence (Yudkowsky, 2003b). SL4s interested in SL4 concepts generally require a strong scientific background to understand these ideas and communicate them to others (this alone doesn't necessarily say anything about their validity, but it remains a fact). Fiction authors John Wright and Vernor Vinge are probably borderline SL4. Non-fiction writers Eliezer Yudkowsky, Nick Bostrom, and myself are all SL4. There are somewhere between 20 and 100 SL4ers out there, depending on what you call "SL4". "Staring Into the Singularity" is probably the first and most intense of SL4 writings online. SL4 ideas and people are apparently so eccentric and unusual that the San Francisco Chronicle, Wired, and Slashdot have all bothered to take the time to make fun of them. SL4s ideas can shock SL2s and even SL3s. SL4s believe that the creation of qualitatively smarter-than-human intelligence could result in discontinuous levels of progress. The comparative value of various SL4 concepts has been an intense topic of debate among transhumanist communities for almost half a decade now, and this will probably continue to be the case indefinitely.

Bibliography:

Bostrom, Nick. (2003). "Ethical Issues in Advanced Artificial Intelligence". Cognitive, Emotive and Ethical Aspects of Decision Making in Humans and in Artificial Intelligence, Vol. 2, ed. I. Smit et al., Int. Institute of Advanced Studies in Systems Research and Cybernetics, 2003, pp. 12-17. http://www.nickbostrom.com/ethics/ai.html

Cosmides, L., Tooby, J. (2001.) Evolutionary Psychology Primer. Webpage. http://www.psych.ucsb.edu/research/cep/primer.html

McKie, Robin. (2002.) Is human evolution finally over? The Observer International. http://observer.guardian.co.uk/

Slovic, P., Finucane, M., Peters, E., MacGregor, D. (2002.) When Predictions Fail: The Dilemma of Unrealistic Optimism. In Heuristics and Biases (pp. 335-347), T. Gilovich, D. Griffin, D. Kahneman (Eds.). Cambridge, U.K: Cambridge University Press.

Smart, John. (2001.) What is the Singularity? http://singularitywatch.com

Koehler, D., Brenner, L., Griffin, D. (2002.) The Calibration of Expert Judgement: Heuristics and Biases Beyond the Laboratory. In Heuristics and Biases (pp. 6-420), T. Gilovich, D. Griffin, D. Kahneman (Eds.). Cambridge, U.K: Cambridge University Press.

Klawans, Harold. (2003). Strange Behavior: Tales of Evolutionary Neurology. New York: W.W. Norton and Co.

Yudkowsky, Eliezer. (1999). "Future Shock Levels". Webpage. http://www.yudkowsky.net/shocklevels.html

Yudkowsky, Eliezer. (2003a). "Predicting the Future". Talk given at the World Futures Symposium 2003, sponsored by the New York Transhumanist Association. http://www.sl4.org/bin/wiki.pl?PredictingTheFuture

Yudkowsky, Eliezer. (2003b). "Levels of Organization in General Intelligence". Publication of the Singularity Institute: http://singinst.org/LOGI/ To appear in Advances in Artificial General Intelligence, Goertzel and Pennachin, eds.