Singularity Watch
HomeNewsletterReading GroupsConferencesPublications"Singularity Studies" LinksDegree ProgramsCritiques

 

Acceleration and Punctuated Equilibrium in Technological Development


© 2005-2008, John Smart. Reproduction, review and quotation encouraged with attribution.

Outline

Introduction

 

Interval Time Compression (Between Major Punctuations)

 

The Intrinsic Smoothness of the Dark Ages

 

Introduction

When viewed from a macroscopic perspective, a statistical case can be made that our planet's average level of physical computation, from a network perspective, smoothly and continually accelerates, in general. Yet the historical record also suggests that our own history of technological development has shown a type of "punctuated equilibrium" in its emergence.

We can roughly point out prominent plateaus (a.k.a., "equilibria"), or long periods of apparently slower growth and technological refinement, where computational acceleration continues only "under the hood" of the system, interspersed between briefer periods of obvious new emergence. These latter periods are characterized by the advent of new technologies ("punctuations") that have broadly disruptive and generally valuable effects on human society.

The prolific systems theorist Kenneth Boulding, for example, recognized that the scope of perceived technological change in the 1890s and 1920s probably seemed appreciably faster, to the average citizen, than in the 1970's, as those earlier eras involved massively disruptive sociotechnological innovations like the electric dynamo, telegraph, telephone, automobile, radio, assembly line, etc. Internet archivist Brewster Kahle has recently made a similar point. Human-observed technological acceleration was probably the fastest circa 1900, with today being somewhat less rapidly changing, in outward appearance.

But we know this doesn't tell the whole story. First, in terms of information flows and technological capacity, not human observation, technological evolutionary development has run exponentially faster every year forward. This has been measured in all manner of informetrics. The more complex the network becomes, the more this acceleration simply dives "under the hood," (think of a modern car engine, stunningly complex) and becomes unseen to the average person. We call this a "cognitive singularity," as we will discuss later. This unseen acceleration in turn leads to periodic punctuations in observed complexity, interspersed by long periods of equilibrium, from the human perspective. Speaking personally, I expect the next major punctuation of equivalence to the 1900's, or to the emergence of the transistor in 1948, will probably not arrive for at least two more decades (more on this shortly).

Second, we know that humans seek a measure of serenity, and they rapidly adjust psychologically to observed changes in their environment. The more sophisticated the technology, the subtler the apparent change becomes. Again, the more change the system can tuck "under the hood" from the perspective of the observer. These are two reasons why Boulding's observation, though likely correct, is irrelevant to the nature and timing of the various human surpassing "singularities" in store for us in coming decades.

In a closer examination of technological punctuations, at least two additional important observations should now be made.

 


Interval Time Compression (Between Major Punctuations)

First, consider the "interval time compression" relative to accelerating change. Each new period of equilibrium, a phase of steady refinement but little dramatic innovation, has become shorter and shorter in time before the next major technological punctuation. The emergence of collective rock throwing, initiated by slow-moving, soft-skinned, large-brained, hominids against far more physically powerful attacking predators. This stunning event is a strong candidate for Earth's first major use of nonbiological technology, as well as the most fundamental reason for human civilization. Ever since, each new interval between punctuations, however we choose to define them, has been dramatically shorter than the last. Specific interval choice is always arbitrary, but the compression is apparently not. A few examples:

3 to 4 million years ago collective rock throwing
1.5 million years ago lever, wedge, inclined plane
500,000 years ago control of fire
50,000 years ago bow and arrow; fine tools
5,000 years ago wheel and axle; sail
500 years ago printing press with movable type; rifle
50 years ago the transistor; digital computers

How close are we to the next major punctuation? Present analysis suggests it is farther away than you might think. The emergence of the commercial internet was clearly a recent punctuation, but this technology has only mildly transformed society to date, far less than many of us in the dot.com era might hope. It will very likely take several more decades for this technology to mature to the point where it will engender the next major event, the advent of a broadly useful commercial conversational user interface, or CUI, an emergence that is expected circa 2020 by several estimates.

Other promising fields all appear to be too young (in the case of nanotechnology and artificial intelligence) or too fundamentally constrained (in the case of biotechnology and neurotechnology) to create comparably significant breakthroughs prior to 2020. Even such continually surprising, hard-to-predict fields as solid state physics seem unlikely, statistically speaking, to create a revolution in this timeframe (while wildcards are always possible, room temperature superconductivity, for example, has a low probability of emerging, and applied physics has been more evolutionary than revolutionary for the last several decades). Moreover, progress in all these fields is critically dependent on the intrinsic development rate of infotechnology (information technology), and short of any major advance in theoretical or applied computer science prior to the CUI, it is likely that its gradual emergence will represent the next major I.A. (intelligence amplification) event, or punctuation, for human culture.

From a computation-centric perspective, the most important "microrevolutions" of the latter half of the 20th century have arguably been the commercial digital computer (circa 1950), the minicomputer (circa 1970), the personal computer (circa 1980), and the networked personal computer (circa 1990). This would argue that another microrevolution can be expected circa 2010. Perhaps the best candidate for that would be ubiquitous 3G cellular phones/PDAs and the weak, location-based and proto-CUI services they will commercialize at that time.

But in a "macro" analysis, the last major revolution was arguably way back in 1948, with the invention of the transistor, and everything that has occurred since has been driven primarily by increasing the use of that amazing device for networking of human brains and automating simple human processes. I would argue that we can expect this paradigm to hold at least until 2020-2030, when the CUI has developed such a degree of utility and affordability that all human beings able to access it at that time will consider themselves to be permanently linguistically connected to electronic intelligence. At that point, global access for the "Bottom Three Billion" will begin to become an international imperative, and the rate of development of CUI algorithms, not transistors, will become the new "fundamental primitive" of our planet's technological evolutionary developmental dynamic.

That latter analysis, if true, would situate us currently two-thirds of the way through a 70 year year macroequilibrium period, between the mass use of transistors and the mass use of a commercial global CUI internet. Micropunctuations aside, we may expect to remain in flatland for a while yet. Unfortunately, even our best efforts don't eliminate the equilibrium periods, but we certainly can help them to shrink dramatically with time.

 


The Intrinsic Smoothness of the Dark Ages

Second, consider the "intrinsic smoothness" relative to accelerating change. Human psychology likes to see and create intervals as mental categories, yet there also remains an intrinsic smoothness underlying computational change in any evolutionary developmental system, be it cosmic, chemical, genetic, memetic, or technologic, when observed from a longer term perspective.

It is sometimes mistakenly alleged, for example, that the West's "Dark Ages" (roughly, from 400 AD to 1400 AD spanning the Fall of the Roman Empire to Medieval Europe until the Black Plague) were a period of global technological stagnation. Yet any good history of technology, Eastern, Western, or New World, demonstrates that while civilizations faltered, our technical advances (increasingly independent of human dramas) continued to broadly and smoothly accelerate during this era.

The most obvious patterns during this period are the cyclic rise and fall of empires, both in Europe and other cultures (e.g., Chinese, Arab, Indian, Southeast Asian, Central American). But technology scholars as broadly distributed as Robert O'Brien in the 1960's and Anne-R-J Turgot in 1766 have also noted a smooth and non-cyclic progression of technological innovation during this time, both globally and in the West. This progression appears surprisingly unaffected by social and political cycles in every culture, including the dramatic regression in learning and intellectual freedom which occurred during much of this period in the orthodox Christian West.

Consider the following illustrative innovations from this thousand-year time period, many drawn from the impressive work, The Timetables of History, 1946/91. This list concentrates on first introductions of important technological innovations, rather than mass use, which would be a more valuable indicator (technological diffusion versus innovation) but is harder to compile. Bold innovations are suggested as perhaps particularly noteworthy for stimulating accelerating change, though this judgement is almost entirely arbitrary and context-dependent. The list also includes a few sociotechnological innovations (e.g. select university, city, and civilization formations) to provide context and scope:

Technological or Sociotechnological Innovation Date (A.D.), Location
Alchemy (pre-science) develops a wide following 410, Europe
Constantinople University 425, Turkey
Powers and Roots (Arybhata) 476, India
Heavy plow; horse shoes; practical horse harness 500, Europe
Wooden coffins (Alemanni) 507, Germany
Draw looms (silk weaving) 550, Egypt
Decimal reckoning 595, India
Canterbury Monastery/University 598, England
Book printing 600, China
Suan-Ching (Science Encyclopedia) 619, China
Originum Etymologiarum Liibri XX (Science Encyc.) 622, Spain
First surgical procedures 650, India
Water wheel (Benedictine monk-engineers) 700, Europe
Stirrup arrives in Europe from China 710, Europe
Early Chemistry (Abu Masa Dshaffar) 720, Mid-East
Medicine, Astronomy, Math, Optics, Chemistry 750, Arab Spain
Hanlin Academy 750, China
Pictorial Book Printing 765, Japan
Iron and smithing become common; felling ax 770, Europe
Chemistry (Jabir) 782, Mid-East
Mayan Acropoli (peak) 800, Mexico
Algebra (Muhammed al Chwarazmi) 810, Persia
Ptolemaic Astronomy; Soap becomes common 828, Europe
Rotary grindstone to sharpen iron 834, Europe
Paper money 845, China
Salerno University 850, Italy
Smithing becomes more complex; Trebuchets 850, Europe
Astrolabe (navigation) 850, Mid-East
Angkor Thom (city) 860, Cambodia
New Mathematics and Science (Jahiz, Al-Kindi) 870, Mid-East
Viking shipbuilding; Iron becomes a commodity 900, Europe
Paper arrives in Arab world 900, Egypt
Salerno Medical School 900, Italy
Linens and woolens 942, Flanders
First European bridges 963, England
Arithmetical notation brought to Europe by Arabs 975, Europe
1,000 volume encyclopedia 978, China
First Mayan and Tiuanaco Civilizations 1000, South America
Horizontal loom 1000, Europe
Astrolabe arrives in Europe 1050, Europe
Greek medicine arrives in Europe (Constantine) 1070, Europe
Water-driven mechanical clock 1090, China
Antidotarum (2650 medical prescriptions) 1098, Italy
Bologna University 1119, Italy
Mariner's compass 1125, Europe
Town charters granted (protecting commerce) 1132, France
Al-Idrisi's "Geography" 1154, Italy
Oxford University 1167, England
Vertical sail windmills 1180, Europe
Glass mirrors 1180, England
Second Mayan Civilization 1190, Central America
Cambridge University 1200, England
Arabic numerals in Europe (Leonardo Fibonacci) 1202, England
Tiled roofs 1212, England
Cotton manufacture 1225, Spain
Coal mining 1233, England
Roger Bacon, our first scientist (Opus; Communia) 1250, England
Goose quill writing pen 1250, Italy
The inquisition begins using instruments of torture 1252, Spain
Tradesman guilds engage in street fighting over turf 1267, England
Toll roads 1269, England
Human dissection 1275, England
Wood block printing; spectacles 1290, Italy
Standardization of distance measures (yard, acre) 1305, England
Use of gunpowder for firearms (Berthold Schwarz) 1313, Germany
Sawmill; wheelbarrow; cannon (large and hand) 1325, Europe
Pisa and Grenoble Universities; Queens College 1330, Europe
First scientific weather forecasts (William Merlee) 1337, England
Mechanical clock reaches Europe 1354, France
Blast furnaces; cast iron spreads rapidly across Europe 1360, Europe
Steel crossbow first used in war 1370, Europe
Vienna, Hiedelberg, and Cologne Universities 1380, Europe
Incorporation (of the Fishmonger's Company) 1384, England
Johann Gutenberg, inventor of mass printing, born 1396, Germany

As we consider this impressive set of emergences, occurring at the same time as the longest era of political and ideological repression in known Western history, we are beginning to learn that technological acceleration is something larger, smoother, and far more powerful than our cyclic biological affairs. Technological innovation appears to be a universally permitted developmental process, one largely independent of wars, enlightenments, reformations, inquisitions, crusades, subjugations, and other aspects of our fluctuating human ideological, cultural, and economic history. As Robert Wright has observed (Nonzero, 2001), technological innovation is something we choose, often unconsciously, regardless of who is in on the throne or in the sanctuary, because it has strong "non-zero sum" effects on human aspirations.

Both of these observations, interval time compression and intrinsic smoothness, can be made in the evolutionary developmental records for cosmic, molecular, genetic, and memetic (cultural) punctuated equilibrium as well. But in the accelerated domain of technology the patterns are far more time-compressed and pronounced, and perhaps the easiest to define for each of these unique physical-computational systems, or "substrates."

Curiously, when we measure technological change closely, using a wide variety of metrics (e.g., data inputs, processing, information generation, connectivity, storage, replication rates, diffusion rates), we discover that it has continually and consistently accelerated over the known history of civilization. Many careful thinkers now suspect that our increasingly self-directing and self-stabilizing technological and computational systems will continue to speed up for the forseeable future, apparently in the same manner that general developmental change has done for the last six billion years, grossly, for the entire second half, of universal history. Something very curious, and potentially universally significant, is going on.

Feedback? Fixes? Critiques? Email johnsmart{at}accelerating.org. Thank you.