We don’t usually lose technologies. In the annals of the technium there are extremely few cases where a society has given up a working technology without substituting something superior. Occasionally specific know-how may disappear. In the 6th century Byzantine empire some pyrotechnical wizards invented a incendiary weapon called Greek Fire, or warfire. This fire could not be extinguished by water, in fact it would burn on top of water, and it would eat itself while incinerating anything it touched, a most miraculous and handy weapon at that time. From a distance of up to 50 feet soldiers would shoot the spectacular fire onto ships and forts, which spread fiercely, and quickly consumed targets with unquenchable flames. But warfire’s exact chemical composition, always a highly guarded military secret shared by a few keepers in the Asia Minor, was lost during the medieval ages. The current consensus is that warfire was crude oil shot through a long bronze pump, and alternatively lobed by a catapult in clay jars wrapped in burning cloths. Perhaps because warfire required crude oil seep wells found only in the mid-east, this technology died out as the most advance warfare technology drifted to Europe.

Another three well-known examples of lost technology: Around 100 BC tinkerers in ancient Greece, maybe even Archimedes himself, invented a “laptop” astronomical calculator, now called an antikythera, that could predict eclipses and serve as a kind of portable planetarium to determine the positions of celestial bodies. There were at least four models of this small analog computer made in ancient times,  but all were lost with no influence on subsequent technology. Two thousand years later scuba divers found a corroded remains in an ancient shipwreck at the bottom of the sea in Greece. The precision, miniaturization, and ingenuity of its metal parts and fine gears of the original calculator were not equaled again until 18th century clocks, yet somehow this sophisticated technology was abandoned for cruder technology.

In the 10th century Arabs compiled a nine-volume encyclopedia of medical knowledge and philosophical notes written by the renowned physician Al Razi, said to be longest medical text ever written by a single person. Two hundred years later this book was forgotten in the Islamic world. No copies in Arabic have ever been found. We only know about it through a Latin translation made in 1279. While some of its wisdom was found in other writings, the bulk of it disappeared in the Islamic world, the center of civilization at the time.

Japan gave up gun technology for 200 years. The first guns in Japan were imported by Portuguese traders in 1543. Within 50 years Japanese master metal smiths had cloned them and were exporting firearms for cash. Even before 1600 guns and cannons were deciding battles in Japan. In a extremely chivalrous culture (10% of Japanese men were samurai) the gun was assigned to low-class uneducated foot soldiers, while the art of the sword was the more honorable choice for rulers. As the gun became common it became unfashionable, vulgar, despicable in its ability to kill indiscriminately. Its slow removal was a type of arms control — the gun was seen as too deadly for civilized people. Europe made similar, but ineffectual, attempts to ban the gun around the same time. Martin Luther called firearms “devilish.” King Henry IV in France prohibited non-royal manufacture of gunpowder and King Henry VIII in England outlawed firearm ownership by anyone earning less than 100 pounds per year. Neither edict was successfully enforced. However, Japan was able to centralize gun production and banish the weapon from their island by 1668. The Japanese reverted to their beloved swords, and since they were closed to outside influence during this period and only fought among themselves, the gun did not return until Commodore Perry forced it in 1879. For two centuries the Japanese retreated from the manufacture of complex firearms and instead perfected the manufacture of the world’s most sophisticated swords (they could slice through a barrel of machine gun).

These four examples of reversal are fascinating primarily because they are so scarce. Despite these exceedingly rare retreats, the technium moves relentlessly forward. Generally, a society does not abandon a new technology to return to an earlier version. When a current technology is suspended in the natural course of evolution it is usually displaced by a more complex variation, and the old version is swept aside as a viable minor alternative, or at least a curiosity, but rarely goes extinct. For instance in the age of automation, older hand tools are perfect for working off the grid, or in tight spots, or in countries with little cash. In an urban world, swords are hammered out by blacksmiths for ritual purposes. Quilts are sewn by hand for recreation and community. Fish are caught by hook for sport. Leather is used for the best shoes because the improvements on leather aren’t really better. Commonly, the transition to the new appears faster than it is, as the old lingers invisibly behind the glittering flash of the new. For instance, despite the dominance of automobiles on modern culture today, more bicycles are sold each year than cars.

Rather than a series of linear displacements climbing a ladder of evolution, the technium progresses as a widening field of accumulation. Existing technologies keep operating almost intact, but are subsumed under additional new, more complex layers. Beneath the fancy hi-tech digital world a viable industrial foundation provides rare-earth metals, electricity, copper wire, and plastic keyboards. Beneath the industrial foundation a viable agriculture swells. As any modern farmer will tell you, the glories of virtual worlds and e-commerce depend upon a rather primitive cycle of poking seeds into dirt and harvesting the replicants into large bins, a routine that has not changed much in 8,000 years. The agriculture era has not disappeared.

Because human population is still growing, the size of the agricultural layer of the technium continues to spread. The shrinking cohort of farmers in the developed world is more than offset by their growing numbers elsewhere. There is more agriculture on this planet than ever before (one third of land surface). In the past 40 years of the industrial revolution alone, cropland has increased globally by 12%. At the same time, and for the same reasons, the widening base of farms and foodstuffs permits industry to expand. The industrial age is nowhere near ending. Its continual expansion permits new post-industrial technologies to expand. The leading edge of technology (lightweight, disembodied, highly leveraged stuff — solar panels, gene therapies, and quantum computers) races forward, but only because its subsumed foundations also march forward.

But why does the technium so rarely go backwards? Why are forgotten calculators, weapons, and medical encyclopedia so uncommon? Why is there a one-way directionality to technical progress, so that in its broadest outlines it inexorably moves towards the more complex with so little retreat?

Four reasons. First, the natural momentum won by any entrenched system bestows a directionality. The more established a process is, the harder it is to change, the more it proceeds along its path. Big technology is hard to stop. For instance, the US electrical system achieved technological momentum by the time it set its standards for voltages and frequencies in 1900, and they have not changed since. For all practical purposes the flexibility of a technological system is eliminated once its initial choices and defaults are fixed. As systems scale up they acquire inertia. The components of global technologies like container freight shipping, or integrated circuit design, or hospital x-ray equipment, tend to create a kind of autonomous movement simply because of their size. 

The scale of many technologies dwarfs us individually. The span of pulsating power lines continue beyond our vision to wrap around the earth.  This grid, built 100 years ago, lighted your grandparent’s home, and our parents’, and now brightens mine, and will light the lights of our grandchildren and probably their grandkids. The wires, towers, circuit breakers, and network of connections might eventually persevere as long as the aqueducts of Segovia, Spain built in 100 CE. That technology conducted drinking water for 75 generations. Roman roads outlived the Romans. This technological longevity is almost a kind of immortality that transcends our comparatively brief lives. The technium’s scope exists outside of our oversight, especially outside of our personal oversight. Its omnipresence together with its relative immortality grants it a version of autonomy.

The second way that technology gains a measure of autonomy is through its self-creation of needs. When we purchase a mobile phone that is just the beginning. The cell phone “needs” a voicemail box, a charger, an ear piece. Previously phones did not need apps but now phones are defined by their cousins, computers, so they do.  As historian David Nye observes, the mobile phone and by extension any tool, “enters into the determination of its own utilities, suggesting new ideas for its own definition.”  The more ways a technology insinuates itself into the existing fabric of the technium, the more autonomy it can acquire.  As its supporting web is cast wider, it can come to rely upon hundreds of technologies, and in turn sustain hundreds itself. And of course each of those assisting technologies is subsidized by a web of others devices. In many cases inventions co-support each other. The entire ecosystem of interrelated technologies begins to act as a whole.

In that curious way of life, growth triggers more growth. The web of technologies is ever expanding because a particular technology will self-generate new needs, new demands, and new appetites. David Landes summarized the technium’s emergent self-making with this succinct story in Unbound Prometheus, his history of the industrial revolution:  “The invention and diffusion of machinery in the textile manufacture and other industries created a new demand for energy, hence for coal and steam engines; and these engines, and the machines themselves, had a voracious appetite for iron, which called for further coal and power.  Steam also made possible the factory city, which used unheard-of quantities of iron (hence coal) in its many-storied mills and its water and sewage systems…  And all of these products–iron, textiles, chemicals–depended on large-scale movements of goods on land and on sea, from the sources of the raw materials into the factories and out again to near and distant markets.  The opportunity thus created and the possibilities of the new technology combined to produce the railroad and the steamship, which of course added to the demand for iron and fuel while expanding the market for factory products.  And so on, in ever widening circles.”

These multiplying connections and ramifications thicken into a whole that exhibits idiosyncratic quirks, and expresses its own needs. We are all familiar with the temperamental gasoline engine that won’t start unless kicked, or the laptop that doesn’t like to go to sleep. A few years ago MIT graduate student Eric Brende dropped out of technological society to imitate the Amish. With each power tool he discarded he came to see machines as autonomous children: “A modern technology machine is no mere tool. It is a complex fuel-consuming being with needs of its own. It gobbles up energy, it demands care and maintenance; it even has bouts of temperament. In many cases no diaper will contain its mess. For these reasons, it not only serves but must be served. But it is more than another mouth to feed; as it becomes more involved and involving, it can easily invade the living space we formerly reserved for ourselves, taking on functions once our own.”

Autonomy in technology is encouraged in a third way: we design it in. Science fiction guru Isaac Asimov offers a good analogy: “A chipped pebble is almost part of the hand it never leaves. A thrown spear declares a sort of independence the moment it is released… The whole trend in technology has been to devise machines that are less and less under direct control and more and more seem to have the beginning of a will of their own.” Clever engineers have designed the calculator that turns itself off, and the rice cooker that turns itself on. Anything “smart” has a bit of autonomy designed into it. Our smart modern kitchens are remodeled with semi-autonomous appliances that decide on their own how much to wash dishes, or toast bread, or make ice cubes. Technicians first engineered autonomy into transportation by creating “auto mobility” in the first automobiles; now cars have the semi-autonomy of cruise control which drives at a steady speed. The next escalation in autonomy is a car that can drive itself. Experimental vehicles such as Stanford University’s  Stanley can navigate itself across country without a human driver. You can also buy self-parking car prototypes. Unmanned auto-piloted airplanes now form combat units in the US Air Force. Asimov’s thrown spear declaring its own independence is now a launched missile that completely steers itself as a bird might.  If this is not a full measure of autonomy, what is?

But the fourth reason autonomy arises in technological systems is the most important. Every technology (even an early Sapien’s hammer) is assembled of parts. When the web of parts contributing to a technology reaches a sufficient degree of complexity, the whole can spontaneously self-organize. Out of this self-organization springs autonomy.

We know from thousands of controlled experiments in many fields of science that complex systems can reliably self-organize. Not in every case, but under certain conditions, stable forms will emerge from messy networks. We’ve seen this emergent order happen in chemical systems, ecosystems, complicated communication networks, biological processes, in human social webs, celestial dust clouds, in evolution and in tens of thousands of computer simulations. “Order out of nothing” emerges when outputs from one  process are diverted as inputs into another, and then circulated back as outputs in an amplifying loop. As an example, anaerobic algae in a warm pond may breed furiously producing a toxin which kills off competing water plants, depleting the oxygen plants normally make, which boosts the growth of anaerobic algae, which kills more plants and so on encourage algae growth, as if the algae had organized the pond. The results of these “strange loops” and recursive circuitry in many complex systems — including technical systems such as computers and the internet — is that a higher level order materializes out of the disorder among many component parts.

The examples of self-organization are legion, and many books have been written about this paradoxical phenomenon, including a very long book that I wrote 15 years ago. At that time I could find no better example of emergent self-organization than a bee hive. Without any central command, thirty thousand individual honeybees can select a new home, manufacture a palace of precise comb for a queen they choose, and oversee the colony’s surplus fuel storage (honey) —  tasks that the bees are incapable of doing individually. Critically, individual bees need the colony to reproduce since all honeybees, except the queen, are sterile. The bee colony is an emergent self-organized form. A beehive can remember nectar locations longer than any bee can, it will outlive the short life span of individual bees, and as a colony it is a warm-blooded animal maintaining a constant temperature while the tiny bees are cold-blooded. Yet, no where in the outward behavior or internal structure of a bee do we see the form of a hive. Where do the plans for the beehive hide? The order of the colony only emerges from the complex interactions between the honeybees. Its form is not imposed from without, or mimicked from the parts, or dictated by the beehive brain, or the queen, but rather the hive is organized by the emergent autonomous “self” of the whole.

Autonomous self-organization can occur in relatively simple physical systems like a sand dune (self-organization is what gives the dune its shape), in the dendritic pattern of a draining river in a delta, in a boiling pot of water (creating tiny vortexes), or in bits of iron that spontaneously magnetized. Specific solutions of chemicals can form themselves into tubular structures (micelles), or into clock-like oscillating reactions, or even self-assemble into more complicated compounds. Life abounds with self-organized patterns such as the leopard’s spots, the spiral coil of snail shells, and the geodesic spheres of diatoms. We know these are patterns that direct themselves and not patterns dictated by the organism itself  because the same pattern can be replicated using random inputs. In other words, if you could take a million blank white zebras and supply their skins with a few simple rules governing how to darken their hides as camouflage, the familiar solution of black and white stripes would appear over and over again.  Rather than being designed by the zebra’s brain, or even governed by their genes, the stripes self-organize from the uncoordinated, decentralized actions of millions of pigment cells in the skin interacting together according to a small set of conditions.

Real zebra left, self-organized model right.

Similar autonomous patterns emerge on the patterns of mollusk shells, and in the bark of trees, and in colonies of fungi. Beautiful dynamic patterns reminiscent of microscopic creatures autonomously emerge in a simple mathematical visualization (called a cellular automata), completely independent of any physical form, strongly suggesting  that these kind of emergent patterns are inherent in any kind of complex adaptive system — including the technium.

There is only one technium on Earth, and it does not lend itself it to easy experiments. But we can demonstrate the technium’s inherent tendency towards autonomous self-organization by modeling the system in computers. At the Santa Fe Institute economist Brian Arthur crafted such a model of simplified technium. His model is based on the idea that in the real world different technologies serve as inputs or outflows for dozens of other technologies. A large factory drillpress might be used to manufacture many smaller retail tools. Your laptop relies on the input of your phone line. The engine in your car needs a water pump which in turn supports a radiator which in turn cools the engine. Each machine might prop up hundreds of others. Recall all the technology that goes down when the electricity goes off!  In Arthur’s simulation, each so-called “machine” is a bit of elemental computer logic. Alone, one scrap of logic can do nothing, but when these elements of logic are strung together the abstract “machine” will accomplish a task, such as sorting a large pile of numbers. Just as in the real technium, tens of thousands of machines in various degrees of interdependencies create a vast dynamic web of relationships. So in the model technium one logic machine might spawn ten other logic machines. It is impossible to unravel the full consequences of removing or adding just one technology, but as in the experiment with zebra stripes, when Arthur re-runs such simulations multiple times, mixing up the combinations of machines and inventions at random, he finds that stable patterns of complex “machines” emerge again and again. These autonomous forms are not reflections of constituent pieces, but are independent of them because if you remove or add a few, you get the same results. In the artificial technium, the specific arrangements between inventions don’t matter.  The patterns are inevitable in this sense. Just as important, the patterns are autonomously self-generated.

Computer scientist Abbe Mowshowitz reminds us that “to assert that technology has become an autonomous agent of change is not to attribute an occult quality to the growth of modern society which transcends human choice. It simply means that mechanization… creates a foundation for further development along certain lines.”  This is what large systems can do: they develop new organizational modules.

Nobelist Herbert Simon conjured a timeless fable to illustrate this principle. Imagine two old watchmakers assembling a batch of fine gold watches built from 1,000 tiny parts each. One of the watchmakers (call him Tempus) starts with the first gear and keeps adding the next part until the watch is done. If Tempus gets a phone call and puts down his work, the delicate assembly falls apart and he has to start over again. However the other watchmaker (Hora) assembles the watch in subgroups of 10 pieces each. Now if Hora is interrupted and puts down his work he loses no more than one hundredth of his progress. Simon calculated that if there was a one in ten chance that the watchmaker’s next step might be interrupted (since both watchmakers had many loyal customers clamoring for their quality craftsmanship) then it would take Tempus on average 4,000 times as long to make the same watch as Hora.

In Simon’s fable, Hora’s invention of modularity forms a ratchet which prevents his progress from backsliding. Science philosopher Jacob Bronowski calls that ratcheting “stratified stability.” Risky innovations are stabilized by operating as modules. The worst that can happen is that complexity will collapse down to the stratum of the previous stable unit. This modular method of advancement is widely used in modern manufacturing for the same reasons of efficiency. Modularity was the key to the unlocking the power of factory mass production in the 1860s when it was first used to make guns: assemble all the firing mechanisms first, then all the handles, then all the barrel sections, then assemble these interchangeable units into rifles. Today large factories that “manufacture” automobiles don’t make anything; instead factories are the interchange  where hundreds of pre-assembled modules (reduced risk) are brought together as the final step. Engineers like Hora usually design the modularity. But in large-scale complex systems, such as a living organism, or in evolution, or in the technium itself, modularity is an emergent pattern.

Modularity is ubiquitous in life, and essential in evolution. The disbelievers in the fact of biological evolution are right about one thing: the chances of all the little atoms found in the first living organism coming together randomly is nil. Even if they did come together once, they would degrade instantly. For life to succeed many such unlikely molecules would have to randomly assemble many times. Such a coincidence is statistically impossible. That’s true if you calculate the probability of all the parts acting independently. But that is not how life works. Evolution progresses by self-organizing in a modular fashion. It creates complexity by stabilizing and building upon rare combinations when they happen. For example, a pool of pre-biotic compounds might occasionally self-assemble the complicated molecule A, which sadly, is unstable. Yet while A exists it can produce molecule B, which is also short lived, but B degrades into C, which is rare yet fairly stable. But as it happens in this case, C acts as a catalyst for the creation of molecule A! It’s a loop! Because C is stable, it can keep making a steady supply of unstable A and B, which pumps up the quantity of C. Suddenly this self-organized recursive circuit fills the pool with persistent, unlikely complex molecules. The probability of the whole is paradoxically greater than the sum of its improbable parts. These stable auto-catalytic molecules serve as modules for more complex compounds in the same self-generating manner.

In the same way, a million monkeys in a million years will never type out Hamlet if they type only letters. But if their typing system evolves a mechanism that organizes keystrokes to type random words — a new modular unit — then suddenly Hamlet is not so unlikely. If further self-organization created keystrokes for random sentences — a yet more complex modular unit — then over a million years Hamlet is almost certain. The progression from modules of letters to words to sentences is stratified stability, and a product of self-organization.

Brian Arthur realized that technology, too, got its traction from the ratcheting buildup of modular units. “New technologies are never created from nothing.” Arthur observed. “They are constructed—put together—from components that previously exist; and in turn these new technologies offer themselves as possible components—building blocks—for the construction of further new technologies.” A modern jet engine is composed of a dozen extremely complex subunits like compressors, turbines, and combustion chambers, which themselves are composed of dozens of smaller, equally complex components. Some of these modules have not changed in function since the first jet engines a century ago. New technologies, Arthur points out, are mostly “fresh combinations of what already exists.” Successful older combinations of simple technologies become foundational modules that are reused, recycled, and retained.  The combination of an electric motor plus pulley was invented shortly after electricity was tamed, but can be found in your car today.

The technium today is entirely populated with combinations of primitive technologies that have been ratcheted up into more complex devices. A century ago scientists combined existing modular components such as the triode vacuum tube and paper foil capacitor into the first amplifier circuit. The amplifier then became a new module that was combined with resistors, ear phones, and other components (themselves assembled from many parts) to produce a radio receiver. The radio found hundreds of uses; combined with other modular ingredients it made a radar; the radar unit, mixed with thousands of other subassemblies, yielded a battleship.

Arthur’s artificial technium was able to invent incredibly complex logic circuits capable of adding 8 bits, a significant feat even for a human programmer.  “Complicated circuits constructed themselves from simpler ones,” Arthur writes.  But these complex inventions came about only because the system created stable intermediate modules along the way. When it discovered a particularly useful logic circuit, like a 2-bit adder, it would  then recombine them to produce more complex forms. Arthur calculated that odds of all 9 varieties of 16 logic units assembling at once in the correct combination to a produce working 8-bit adder circuit was one in 10^177,554 — or a certain impossibility. Yet here the ‘miraculous” machine was, operating in his tiny world.  The ratcheting in his artificial technium produced improbable novelty.

In the technium, as in other complex systems, self organization regularly yields improbable novelty,  a ratcheting up of levels, a rise in autonomy, and an unsettling sense of the inevitable. Over time, self-organizing also generates a non-mystical directionality. This last result may take a bit of explaining:

A long arc flows through the cosmos. It begins with the first bits of matter, then is lifted by life, and passes through the technium today. Presumably it continues into the future. The course and direction of this great arc began long ago in the physics of the big bang. Most of universe then, as well as today, operates in the realm of the very very tiny, and includes most of what we consider real: atoms, light, radiation, chemistry, gravity.  With few exceptions the laws governing this microworld are reversible. A time-lapse movie of atomic particles colliding into each other, or zooming around space, would be indistinguishable played backwards or forward. This physical reversibility at the smallest scale is an essential trait of the material world. In a formal sense, time does not have a direction in this realm.

Time only gains direction in the macroworld, the scale at which we live. In the realm of planets, stars, volcanoes, periwinkles, elephants, and of course movies, we see an ineradicable direction, which we call time. However it is not the larger scale of the macroworld per se which points time one way. Sensual time gets its direction from extropy. Extropy (also known in physics as negentropy, or negative entropy) is the cosmic force behind the great arc that assembles complexity into larger and larger visible scale, and that builds up the universe while entropy drains it down. The contrast between the two — entropy and extropy — gives us our embodied sense of time.

Narrow window of feasibility (white area) in trajectories of expanding Universe. (From Rees)

The seeds of extropy are infused into the particulars of the micro world. Ours is a very precise universe. The uniformity of motion and mass from one end of the universe to the other is astounding. Furthermore, if the strength of the forces between parts of the universe, the invisible energies binding its myriad particles together, had varied even infinitesimally from the values it does have, the universe (as we know it) would not hold together. These exact values hold a surprise. The precise settings of the 30 most important constants in the universe — like the value of gravity, the energy density, the strength of electromagnetism — enable bits of matter and energy to organize themselves into structures which can not only persist in the face of entropy, but persist with modification.  Extropy is the innate inclination buried in the very fundamentals of reality for small things — starting at the subatomic level — to cooperate as a larger unit (could be an electron, an atom, molecule, or a cell) long enough to extend the inclination toward yet larger units. It is a self-ballooning force that bootstraps its way into greater existence.

The very slow cooperation of subatomic particles to form the hundred or so elements in the furnace of stars took several billion of years. The massive self-organization of stable planets took several more billions of years. Because extropy is a self-creating cycle it ramps up slowly from almost nothing. But the more structure it creates, the faster it creates more structure. Biological evolution accelerated extropy in at least one local area (Earth); technology even more so. The technium is overrun with self-generating powers.

“What has brought life by slow steps up a ladder of increasing complexity is stratified stability,” says Jacob Bronowksi in his Ascent of Man lectures. “And we know this is true not only of life but of matter. If the stars had to build a heavy element like iron, or a super-heavy element like uranium, by the instant assembly of all the parts, it would be virtually impossible. No. A star builds hydrogen to helium; then at another stage in a different star helium is assembled to carbon, to oxygen, to heavy elements; and so step by step up the whole ladder to make the ninety-two elements in nature.” 

The self-organization that is common to chemistry, life, and the technium moves through the universe and time in the same way. From myriad parts extropy organizes structures that lock-in order at a higher level — stratified stability.  The new emergent order stabilizes as a new “whole” which then serves as a component part for the next round of self-organization. So the atoms that build themselves into a glycine molecule are subsumed into a protein; then that protein, with millions of other proteins, is subsumed into the higher order of a whole cell. And so on. The technium, too, is built this way.

It is this stratified stability created by extropy, evolution, and self-organization which prevents the collapse of complexity. If complexity had to re-assemble itself at every instance, nothing really complex would be possible.  Structures as sophisticated as ourselves would be unthinkable. How could our bodies, let alone our minds, arrange themselves from a billion atoms at once? The modular nature of self-organization encourages complexity to bootstrap creations, and it prevents backsliding. Extropy keeps things moving forward. In Bronowski’s lovely phrase, it “gives the arrow of time a barb, which stops it from running backward.”

That’s a gorgeous image.  It’s the irreversibility of extropy that gives us a sense of things going forward, and supplies the “barb of time” that sets our world apart from the reversible microcosm of the quantum world. All the time-lapse sequences that fail to play backwards are sequences generated by extropy: the development of a toddler from an infant, the evolution of a fish from a sea worm, even the slow aggregation of this planet from sun dust. Billions of these parallel narratives, all impelled by the extropic barb of time, smother us with the impression that universe flows in only one direction (even if it is only the things we care about that actually do).

But extropy does even more than illuminate the temporal. The ratchet of its self-organization is the “barb” which pushes progress. Progress is not an mirage. Progress is what evolution and extropy do. Complex systems tend to get more complex because the levels of emergent order keeps systems from devolving in reverse. Order doesn’t retreat (on average) while it remains pushed by extropic system. The stratified stability of evolution and self organization act as a valve, constantly directing change in one direction — towards higher levels. On average biological organisms rarely devolve into simpler organisms. Almost always categories of living creatures evolve towards more complexity, more specialization, more diversity, more sociality, more sentience, and ubiquity. The same is true of individuals. We start off as a single undifferentiated cell and develop into more complex beings. Likewise our institutions, which tend to get more complex in their own lifespans.

The technium too advances through an escalation of stable coherent forms. Once  crops are domesticated  they almost never go wild again. Once a society adopts agriculture it rarely surrenders its farms to return to hunting; most farm societies evolve crafts and more technology. Once money is common, it rarely disappears. Instead coins beget banking, investment, interest. Once electricity is installed in a region, it does not dim. Yes, wars and bandits can rip out its wires in a temporary illness, but in general electrification forms a stable base upon which the complexities of industry stand.

This is why the loss of a technology is so rare. The ratchet of self-organization prevents technological backsliding, and so we marvel at how an ingenious weapon like Greek warfire could have been lost, or how an analog computer built 2,000 years ago could have been forgotten. The particulars are perplexing, but these are the statistically expected exceptions. Evolution is a probabilistic function, and there will be outliers like the retreat on guns in Japan. The real marvel, instead, is the remarkable consistency of cosmic progress.

From the moment of their genesis at the big bang, matter and energy have coalesced into increasingly complicated units — despite the relentless resistance of entropy. Quantum particles clump into subatomic particles, and these in turn clump into lightweight atoms. In the fires of stars 12 billion years ago massive quantities of lightweight atoms combine with more cosmic particles under their own self-generated stellar gravity to generate heavier elements. These heavy elements self-organized into planets, some exhaling self-sustaining atmospheres. All the while entropy nibbles at these islands of self-order. In small pockets on a planet, atoms self-organize a most remarkable molecule: a self-replicating double helix. While self-organization continues to birth emergent order in the cosmos, the double helix unleashes an additional generative force: evolution. Now self-organization is amplified by the learning that evolution brings, and extropy is accelerated. The resilient molecule of DNA creates a stable base for the invention of millions of adaptable organisms, and each of these creatures — every one of the trillions that have lived — create more order in one corner of the universe, and pay the price with more entropy. Through innumerable stages of increased structure (limbs, nerves, eyes), the biosphere continues to arrange itself in greater units of order until it learns to improve its learning with a brain. Biology forms an irreversible base for the launch of mind, which can now generate more complexity, faster, then ever before. The mind adds another force to extropy. It unleashes the technium, which multiples conscious learning, broadcasting its ability to create new levels of complexity and structure completely around the planet. The technium stands upon the base of biology (and therefore must nurture it). Biology stands upon the base of self-organized matter in the swirling galaxies and twirling planets. And all these layers rest upon the precise initial conditions at the big bang that still solicit this billion-years unfolding.

Matter and energy’s propensity to self-organize propels them outward in celestial forms, life, mind and the technium. Self-organization ascends the levels of complexity and bequeaths the great thrust not only its directionality, which gives it an disconcerting sense of inevitability, but also with its increasing autonomy, which also unsettles us.

The circle of my closest friends are fans and boosters of technology. They unleash the technium for a living. Their job is to discover new forms of machines, to invent new ways to leverage intelligence, and to create entirely new stuff. When I asked them about the inherent value in technology, the majority of the technophiles I interviewed claim that technology is a means to an end, and therefore neutral. They generally hold an upscale version of “guns don’t kill, people do.” We have a choice in how technology is applied, but fundamentally, they would say, technology is neither good nor bad, it’s neutral. The good in technology comes from wise decisions in employing it.

To be honest, I used to feel the same way. History counseled that dynamite could be used to carve tunnels or blow up schools. Insecticides could boost crops or poison drinking water. GPS satellites can guide you if you are lost, or track you down with no place to hide. Surely the sum value of new invention was up to us. And the idea that we choose the valence of technology’s charge is very appealing to our egos. But it does not match the evidence of technology’s rise, nor its deep roots in life and the cosmos.

Nothing that re-arranges energy in the face of entropy; that builds up dynamic forms lasting billions of years (galaxies); that is able to bootstrap self-replication from inert molecules (DNA); that stabilizes increasing complexity (life), that crystallizes permanent change in the form of a mind (you), that amplifies learning with tools (technology) — nothing like that can be neutral. Is life neutral? Is your mind neutral? Is a baby neutral? While we might list certain species or ideas as negative, we understand that in general the more life and ideas, the better. The technium is in the same class, only more so. Technology is as neutral as the biosphere, which is to say that it is not neutral. It comes with biases and autonomy.

There are good and bad uses for television, but the media of television itself is biased toward certain uses — and certain abuses. As Jerry Mander argued in Four Arguments for the Elimination of Television, the physical nature of television of the  1960-1970s— its flickering cathode ray screen images broadcast one-way over a centralized network  — induced a bias of physiological passivity. There was nothing to do but sit back and receive. TV as it was structured was biased to producing couch potatoes.  It was also an effective propaganda device. In the west most of the propaganda was generated by corporate advertising. In Mander’s view television manipulates our psyche with clever emotional pitches to convince us we need more technology  in our lives; that we can’t be happy without a cellphone or car. Among the gadgets it persuades us we need are more media devices, which in turn will manufacture more desires for more technology. Thus technology encourages (or forces) us to produce more technology , so we become mere vehicles in technology’s takeover.

That wasn’t television’s only bias, because TV also accelerated the transition to the middle class in developing countries like Brazil and India. The addictive soap operas that seemed to recycle a small set of human dramas, also acted as incredibly effective role models for illiterate women, who decided their daughters should be like the women on TV. More than any government contraception program, the passive consumption of mindless TV persuaded women into having smaller families and educating their daughters. As Jerry Mander pointed out, as a propaganda device, television is unrivaled.

Abbe Mowshowitz, a computer scientist, puts it nicely: “tools insist on being used in particular ways.”  Neil Postman, who wrote says “the uses made of technology are largely determined by the structure of the technology itself.”  He argues that each technology’s architecture suggest the metaphors we use to manage it, its speed biases the politics and the economics it generates. Mander declared that  “Many technologies determine their own use, their own effects, and even the kind of people who control them. We have not yet learned to think of technology as having ideology built into its very form.”  Thus, if devices have ideologies built into them, big technology may require big centralized government.  Alfred Chandler, a die-hard capitalist who wrote a monumental history of American business, argues that the construction and day-to-day operation of the technium’s industrial subunits, such as its sprawling transportation network, skyscraper cites, food and water systems, and communication grid all required large-scale, centralized hierarchical organizations. He says matter-of-factly, “the operational requirements of railroads demanded the creation of the first administrative hierarchies in American business.” Die-hard environmentalist Dennis Hayes takes it to the next step: Because nuclear power relies on many layers of fail-safe security, and heavy doses of police and soldiers to guard it from terrorism, “the increased deployment of nuclear power facilities must lead society toward authoritarianism.  Indeed, safe reliance upon nuclear power as the principal source of energy may be possible only in a totalitarian state.”

Like in the emerging internet today, the emerging electrical grid a century ago was seen as  vehicle for ideology. David Nye reports that Charles Steinmetz, the leading scientist at General Electric in its first decades a century ago, “expected socialism to emerge along with a national electrical grid.” Steinmetz claimed: “The electric power is probably today the most powerful force tending toward coordination, that is cooperation [socialism].” And, “Lenin famously declared that only when the Soviet Union had been completely electrified could it attain full socialism.”

We know that large-scale technologies can sport fundamental biases because they have been deliberately designed to do so. In a famous example, the legendary New York urban planner Robert Moses restricted the public — that is the poor, lower class public — from new city parks and beaches by a clever and subtle technological discrimination. According to his biographer, Moses lowered the clearance on the 200 overpasses he built on the Long Island Parkway so that public buses (the untidy poor) were unable to use this highway to get to Jones beach, but (middle class) cars could. If intentional bias can be inserted so effortlessly, couldn’t inadvertent bias emerge just as easily?

Ironically, those who most clearly see that technology is not neutral are those who recoil from it. The writings of technology critics Neil Postman, Wendell Berry, Jerry Mander, Langdon Winner, Eric Brende, Marshall McLuhan offer keen insights into the autonomous nature of the technium. But they also rail against it.  They worry about it’s momentum. No one has written about the autonomous nature of the technium with more clarity, nor with more alarm, than Jacques Ellul.

Ellul was a French theologian born in 1912 of aristocratic European parents.  He was educated in an Eastern Orthodox school (his grandmother was Serbian) and as a young man he experienced a mystical religious conversion when he was 22. After delving deep into Christian theology, he came under the spell of Karl Marx, and later joined the French Resistant movement against the Nazis, where he worked to rescue Jews. Theologically his ideas breached the usual denominational categories, so he was never fully embraced by any church, sect, or even secular movement. For instance, he preached universal salvation, meaning everyone is saved no matter what they do, a dogma that did not endear him to the orthodoxy. He identified himself as a Christian Anarchist, which meant he belonged to a church of one. Yet he was a prolific theologian, and wrote almost 50 books on Christian theology. But Ellul is best known for his one monumental book on the technium published in 1954, and released in English in 1964 as The Technological Society.

In French Ellul called it la Technique, and he meant what I mean by the technium  — all the hardware, instruments, and cultural inventions, institutions and technological infrastructure that flow from our inventions. For Ellul, the rise of technique, or the technium, was a sacrilege against God. Our collective creation, the technium, had taken on a autonomous life of its own, a life much bigger than our own, and we were beginning to idolize it. We were being bewitched by this most powerful earthly force, and it was our most holy duty to see the technium for what it was — a spiritual “tyrant”  — and resist its dominance. Ted Kaczynski, the disgruntled math teacher who built postal bombs aimed at murdering prominent technologists claimed to have read Ellul’s book five times while hiding out in his remote Montana cabin.

To Ellul the emerging autonomy of the technium was starkly obvious, though in the early 1950s it was not so obvious to others. Throughout  La Technique, Ellul sounds his alarm: “Technique has become autonomous… It is a power endowed with its own peculiar force… a reality in itself, self-sufficient, with its own special laws and is own determinations… It is an end in itself.” “Technique pursues its own course more and more independently of man. This means that man participates less and less actively in technical creation, which, by the automatic combination of prior elements, becomes a kind of fate. Man is reduced to the level of catalyst.”  Ellul argued the rising technological autonomy of the technium lowers the human being to “a slug inserted into a slot machine.”  In fact, Ellul declares, “there can be no human autonomy in the face of technical autonomy.”

Of course, Ellul was not alone. He had compatriots and disciples. Rene Dubos: “Technology cannot theoretically escape from human control, but in practice it is proceeding on an essentially independent course.” John Kenneth Galbraith: “I am led to the conclusion, which I trust others will find persuasive, that we are becoming the servants in thought, as in action, of the machine we have created to serve us.”  Martin Heidegger: “No one can foresee the radical changes to come.  But technological advance will move faster and faster and can never be stopped.  In all areas of his existence, man will be encircled ever more tightly by the forces of technology.  These forces, which everywhere and every minute claim, enchain and drag along, press and impose upon man under the form of some technical contrivance or other–these forces…have moved long since beyond his will and have outgrown his capacity for decision… Technology is in no sense an instrument of man’s making or in his control. It is rather that a phenomenon that is centrally determining all of Western history.”

These technophobes are right about so many things. The technium is the phenomenon that is centrally determining all of Western history. It is outgrowing our capacity of understanding. It is advancing faster and faster without end. It is proceeding on an autonomous course. It is an end in itself. It is a kind of fate.

This is scary. We have birthed a child more powerful than us, rocketing off to remake our essential nature, yet it zooms beyond our capacity to understand or control, accelerating in power, yet biased in its direction. No wonder the autonomy of the technium provokes such genuine concern.

Yet the very same innate forces of extropy and self-organization that nurture the technological imperative, also are responsible for real progress. We have birthed a child more powerful than us, rocketing the advance of diversity and intelligence,  it multiples on its own, yet it is headed in the direction we’d all like to go — more options, choices, possibilities and free-will.

The very best things in the known universe are products of this thrust. Redwoods, the milky way, dolphins, antibiotics, garage door openers, the members of your family, the internet, and the 5 kilogram organ now reading this have all been summoned forth by the process of extropy and self-organization. The real advances in human society in the last 8,000 years — increasing longevity, increasing literacy, increasing knowledge of our home, increasing circle of empathy for others — have all been due to the same essential force that also scares us: the inevitable technological imperative.

This cosmic drive — this measurable, falsifiable, actual, non-mystical organizing principle — creates the mechanical autonomy that worries us, and the sense of inevitable that unsettles us; yet this same cosmic organizing principle also generates the progress we want and will always head towards.  We can’t have one without the other, and this stirs up an irreconcilable tension.

I believe this Janus-faced god is the source of our great unease. We fear the technium for its otherness and lack of control, and yet we crave it for its blessings. We hate it for it usurping our autonomy, yet we love it for increasing our freedoms. We deny its inevitable progression, yet we expect its inevitable progress. The deal seems to be that technology can improve our lot, only if it can improve its own at the same time.

Ellul, I am sure, would state the deal different terms. He might say: The cost of human progress (and let’s grant it is real) is the loss of human autonomy.

Would that be worth it? If humanity never changed, if human nature was fixed, immovable, or sacred it would not be. Sacrificing such a gem would too steep a price for progress. But is human nature taboo? There are many among us who define the current state of humanity as sacred, and worth protecting, and will refuse progress because it will inevitably tamper with our spirits. I am not one of them.

I think that humanity is our own invention, that we are a technology ourselves, particularly as we live in the modern world. Humans are tools disguised as animals. We have long ago ceased to be independent beings, and as a species are symbiotically dependent on the technium. If all technology were to disappear tomorrow — and by that I mean all — stone points, clothes, control of fire, spears included — most of humanity would disappear in a matter of weeks, and the remnant that survived without any tools whatsoever would not be called human by any of us now.

As the technium rises, so does its gift of progress. But as progress rises, so does the technium’s autonomy. As a self-made invention, we are an element of technology’s rise. “To quarrel with technology is to quarrel with the nature of man,” says Jacob Bronowski. Based on our current direction, humans will move closer to symbiotic union with technology each year forward. Because of this deep dependency we will redefine ourselves, just has we have been doing for 8,000 years. The technium’s long arc will widen our circle of identity and I suspect we’ll come to see the technium’s autonomy as part of our own.