Is History
Converging?
Again?

Jürgen
Schmidhuber

June 2006
(updated 2012)


Associate an error bar of not much more than 10 percent with each date below

graphical convergence
This page is based on the concluding section of J. Schmidhuber: New Millennium AI and the Convergence of History. In W. Duch and J. Mandziuk, eds., Challenges to Computational Intelligence, Springer, 2006. Preprint: arxiv.org/abs/cs.AI/0606081, June 2006. Update of 2012: In Singularity Hypotheses, Springer, 2013. PDF of preprint.

Summary. Essential historic developments (that is, the subjects of major chapters in many history textbooks) match a binary scale marking exponentially declining temporal intervals, each half the size of the previous one and equal to a power of 2 times a human lifetime (roughly 80 years - throughout recorded history many individuals have reached this age). It seems that history itself is about to converge around 2040 in an Omega Point (Teilhard de Chardin, 1916) or Historic Singularity (Stanislaw Ulam, 1958) - compare Schmidhuber's computer history speedup law. Or is this impression just a by-product of the way humans allocate memory space to past events? Scroll down for more!

~29 lifetimes or
~40,000 years ago
skull of homo sapiens
Modern humans start colonizing the world from Africa
~28 lifetimes ago or
~20,000 years ago
illustration: arrows from the cover of the book by Noel D Justice
Bow and arrow invented; hunting revolution
~27 lifetimes ago or
~10,000 years ago
Jericho, home of one of the oldest settlements
Invention of agriculture; first permanent settlements; beginnings of civilization; colonization of the Americas
~26 lifetimes ago or
~5,000 years ago
Schoyen collection: Archaic Sumerian on clay, Sumer, ca. 3200 BC
First high civilizations (Sumeria, Egypt), and the most important invention of recorded history, namely, the one that made recorded history possible: writing
~25 lifetimes ago or
~2,500 years ago
Laokoon group (Roman copy, 140 B.C.)
Axial Age: The ancient Greeks invent democracy and found Western science/art/philosophy, from algorithmic procedures and formal proofs to anatomically perfect sculptures, harmonic music, and organized sports. Old Testament written; major Asian religions founded. Persian empire is only empire ever to encompass close to 50% of humankind. High civilizations in China, origin of the first calculation tools, and India, origin of the zero
~24 lifetimes ago
8th century AD
Chinese bookprint: Diamond Sutra, not the world's oldest, but the oldest dated printed book: 868 A.D. (Britisch Museum)
Bookprint (often called the most important invention of the past 2000 years) invented in China. Islamic science and culture start spreading across large parts of the known world (this has frequently been called the most important event between Antiquity and the age of discoveries)
Omega - 23 lifetimes
15th century AD
Gutenberg press, Niedersaechsische Staats- und Universiaetsbibliothek Goettingen
The largest empire ever stretches from Korea all the way to Germany. Chinese fleets and later also European vessels start exploring the world. Gun powder and guns invented in China. Rennaissance and printing press (often called the most influential invention of the past 1000 years) followed by the Reformation in Europe. Begin of the Scientific Revolution
Omega - 22 lifetimes
18th century
From: Steam Engines Familiarly Explained, 1836
Age of enlightenment and rational thought in Europe. Massive progress in mathematics, astronomy, optics, and physics in general; first flying machines; first steam engines to power the Industrial Revolution
Omega - 2 lifetimes
around 1880
the Voelklinger Huette steel mill, a symbol of the second industrial revolution
Combustion engines, cheap electricity, and modern chemistry power the Second Industrial Revolution. Birth of modern medicine through the germ theory of disease. Genetic theory, evolution theory, globalization, Marxism, European colonialism at its short-lived peak
Omega - 1 lifetime
around 1960
Earth and its 20th century population explosion through the Haber-Bosch process
Modern post-World War II society and pop culture emerges. The world-wide super-exponential population explosion is at its peak. Threat of humanity's extinction through hydrogen bombs; first spacecraft and commercial computers; DNA structure unveiled
Omega - 1/2 lifetime

around 2000

Microchip layout
Cheap personal computers and the World Wide Web power the Third Industrial Revolution - at the moment of this writing (2006) we are still living through this. A mathematical theory of optimal universal artificial intelligence emerges - will this be considered a milestone in the future?
Omega - 1/4 lifetime
around 2020
graphical convergence
First superhuman visual pattern recognition results in limited domains. First computers match brains in terms of raw computing power.
Omega - 1/8 lifetime
around 2030
??
Omega - 1/16 lifetime
????
Omega - 1/32 lifetime
????????
Omega - 1 year
????????????????
Previous publications on this topic: According to Schmidhuber's speedup law (2003), since 1623 the delays between successive radical breakthroughs in computer science have decreased exponentially: each new one came roughly twice as fast as the previous one; they seem to converge around Omega = 2040. Compare the original article arxiv:cs.AI/0302012 or the local copy. Also compare the concept of an approaching historic singularity (Stanislaw Ulam, 1958), which apparently inspired Vernor Vinge's writings on the technological singularity. See also more recent books on this topic by Hans Moravec, Ray Kurzweil and others.


Fibonacci Web Design by
Juergen Schmidhuber
Omega!
And then
what?
Wishful thinking? The following disclosure should help the reader to take this list with a grain of salt. The author, who admits being very interested in witnessing the Omega point, was born in 1963, and therefore perhaps should not expect to live long past 2040. This may motivate him to uncover certain historic patterns that fit his desires, while ignoring other patterns that do not.
Again? Others may feel attracted by the same trap. For example, Kurzweil (2005) plots exponential speedups in sequences of historic paradigm shifts identified by various historians, to back up the hypothesis that "the singularity is near." His historians are all contemporary though, presumably being subject to a similar bias. People of past ages might have held quite different views. For example, possibly some historians of the year 1525 felt inclined to predict a convergence of history around 1540, deriving this date from an exponential speedup of recent breakthroughs such as Western bookprint (around 1444), the re-discovery of America (48 years later), the Reformation (again 24 years later - see the pattern?), and other events they deemed important although today they are mostly forgotten.
Speed Prior Logo
The nature of human memory. Could it be that such lists just reflect the human way of allocating memory space to past events? Maybe there is a general rule for both the individual memory of single humans and the collective memory of entire societies and their history books: constant amounts of memory space get allocated to exponentially larger, adjacent time intervals further and further into the past. For example, events that happened between 2 and 4 lifetimes ago get roughly as much memory space as events in the previous interval of twice the size. Presumably only a few "important" memories will survive the necessary compression. Maybe that's why there has never been a shortage of prophets predicting that the end is near - the important events according to one's own view of the past always seem to accelerate exponentially.

A similar plausible type of memory decay allocates O(1/n) memory units to all events older than O(n) unit time intervals. This is reminiscent of a bias governed by a time-reversed Speed Prior.

Juergen Schmidhuber's TEDx Talk at TEDxLausanne 2012: When Creative Machines Overtake Man

20 Jan 2012: TEDx Lausanne: Talk (12:47) on converging history and Omega Point. See Transcript. (All TEDx talks.)

Also compare: J. Schmidhuber. Celebrating 75 years of AI - History and Outlook: the Next 25 Years. In Proc. 50th Anniversary of AI, p. 29-41, LNAI 4850, Springer, 2007. Preprint: arxiv:0798.4311.