by Bart Kosko
NANOTIME
Bart Kosko
© Bart Kosko 1997
Bart Kosko has asserted his rights under the Copyright, Design and Patents Act, 1988, to be identified as the author of this work.
First published in 1997 by Avon Books, Inc.
This edition published in 2019 by Endeavour Venture, an imprint of Endeavour Media Ltd.
For the whirling dervishes
and their ecstatic search for the light
Between the mirror and the heart
is this single difference:
the heart conceals secrets
while the mirror does not.
Jalāl-ud-Dīn ar-Rūmi
(1207–1273)
Table of Contents
Introduction to the Rerelease of Nanotime
Prologue
Chapter 1
Chapter 2
Chapter 3
Chapter 4
Chapter 5
Chapter 6
Chapter 7
Chapter 8
Chapter 9
Chapter 10
Chapter 11
Chapter 12
Chapter 13
Chapter 14
Chapter 15
Chapter 16
Chapter 17
Chapter 18
Chapter 19
Chapter 20
Chapter 21
Chapter 22
Chapter 23
Chapter 24
Chapter 25
Chapter 26
Chapter 27
Chapter 28
Chapter 29
Chapter 30
Chapter 31
Chapter 32
Chapter 33
Chapter 34
Chapter 35
Chapter 36
Chapter 37
Chapter 38
Chapter 39
Chapter 40
Chapter 41
Chapter 42
Chapter 43
Chapter 44
Chapter 45
Chapter 46
Chapter 47
Chapter 48
Chapter 49
Chapter 50
Chapter 51
Chapter 52
Chapter 53
Chapter 54
Chapter 55
Chapter 56
Chapter 57
Chapter 58
Chapter 59
Chapter 60
Chapter 61
Chapter 62
Chapter 63
Chapter 64
Chapter 65
Chapter 66
Chapter 67
Acknowledgements
Selected Bibliography
About the Author
Introduction to the Rerelease of Nanotime
What will the next world war look like?
That world-war question led to the novel Nanotime. Readers of the rerelease of Nanotime may find this background of interest. The novel tells the tale of a selfish young man who makes it to digital heaven by way of World War III.
The world-war question had a slow gestation and a sudden birth.
The question’s slow gestation came out of years of work in artificial intelligence and earlier work in aerospace during the last years of the Cold War.
I was just 23 when I first worked on naval strike planning for the Tomahawk cruise missile while still in graduate school in San Diego. That was the summer of 1983. The next year I proposed something called the Adaptive Distributed Space Battle Management System for the Strategic Defense Initiative (known as “Star Wars”). I was 26 when I proposed to the United States Air Force a new defense program called ADAM: Air Defense Associative Memory. ADAM was a type of distributed neural network in the sky. I based ADAM on my new family of machine-learning algorithms called bidirectional associative memories (BAMs). BAMs remain an active area of research in machine learning. The Air Force funded the first analysis phase of ADAM as part of its Air Defense Initiative (known as “Air Wars”).
Then I gave up my security clearance and left aerospace. I left for academic life at the University of Southern California and full-time work in what we today broadly call artificial intelligence. My textbook Neural Networks and Fuzzy Systems appeared three years later in 1991. That was the same year that the Soviet Union collapsed. The Cold War had officially ended.
The question’s sudden birth came three years later in May of 1994. It came straight out of Hollywood: It came from a phone call from master film director Oliver Stone.
I had gotten to know Oliver Stone after my trade book Fuzzy Thinking appeared the year before. He called now and asked for a film treatment of something to do with artificial intelligence. Stone had heard that another film master had gotten studio approval to make a movie called AI. That film master was Stanley Kubrick. The two auteurs had earlier made iconic and very different films about the Vietnam War: Platoon and Full Metal Jacket.
It was quite the opportunity for a 34-year-old writer awaiting his academic tenure decision. The only fiction I had published at that point were short stories in various magazines.
But a film treatment about what exactly?
Stone had just two suggestions: Make it big and make it something you care about it. That was it. Make it big and make it count. So off I went to the East Sierra to think and fly-fish and think some more.
The big part was easy: What could be bigger than World War III?
World War III may come someday given our ever-improving weapons systems and exploding computing power and given our hominid endocrine systems. Such a war would likely be a smart war that includes cyberwarfare as well as arsenals of smart weapons. It might include tactical or strategic nuclear weapons as the conflict escalates. It might even include micro-nukes if the nano-engineering advances far enough. But a smart war would depend more on information than on chemistry or physics. That reverses an old precept of war: It would then be cheaper to attack than defend. An example in miniature is a computer virus.
The problem was that a world war seemed quite unlikely in those comparatively peaceful days of 1994. The United States had just entered the Megatons to Megawatts program with Russia. That noble 20-year effort saw US nuclear reactors burn hundreds of tons of down-graded uranium from disassembled Russian warheads. The mid-1990s were the pre-9/11 days of the alleged “end of history.” There would be no more invasions or wars. There would instead be only trade and treaties. Nation states would soon all reach their evolutionary endpoints as liberal democracies.
The caring part was even easier: I care about not dying. Not dying counts.
Death loses much of its ancient terror in the age of machine intelligence and molecular engineering. That does not mean that we can avoid death. Some form of death is inevitable because of the long-term effects of entropy or the average trend toward disorder. Disorder conquers all in time. Even the protons in atoms will likely fall apart in about a trillion trillion trillion years. The last black holes will eventually radiate away into white noise as the universe quietly dies.
But death itself need not be irreversible even if it is inevitable. Death remains an engineering problem. It is a stupendous engineering problem but not an insurmountable one.
The task is to fix nature’s great design flaw of the human brain: It has no back-up. Who would design a brain with no backup?
This suggests a way to sidestep the problem of death. We did not learn to fly by growing wings or by making other changes to our bodies. We learned to fly by powering machines to fly and then putting our bodies in those machines. The same strategy should work for backing up our brains in a chip or in some other computing medium. We can put our backed-up brains in computer chips.
Nor is backing-up or emulating a brain science fantasy. The National Academy of Engineering lists reverse-engineerin
g the human brain as one of its goals for the 21st century. You can back-up what you reverse-engineer. You can also improve what you have backed up. A corollary benefit is that dying in a back-upped brain may be no big deal. It might even someday be a form of extreme sport.
Many things change if we shift our minds from wet synapses to digital devices made of silicon or nanotubes or other physical media.
The big change is time: Thinking in a chip would take place billions or more times faster than thinking in meat. Nerves conduct electrical signals at only a few hundred miles per hour. Many electrical devices conduct signals at a decent fraction of the speed of light. Our subjective sense of time would differ in kind in chip consciousness based on signal speed alone. That sense of time deserved its own name. So I came up with the term nanotime to describe it.
The result was a 40-page film treatment.
We never got the movie made although we sure tried. The computer-generated effects alone would have cost too much in 1994. Stone went on to make the film Nixon. Kubrick ended up handing off his AI movie to Steven Spielberg. It appeared as AI in 2001.
The novel Nanotime grew from that 40-page treatment in the summer of 1994.
The next creative shock also came in 1994 but not from Hollywood. It came from academia: I got tenure on the first of August of that year. Getting tenure is always cause for celebration in the professoriate. It is also a good time to write a book.
So off I went scuba diving in the British Virgin Islands to celebrate. I just happened to take with me philosopher John Stuart Mill’s Autobiography. I spent the mornings diving and then later in the day reading Mill and walking the cactus-covered island of Virgin Gorda. I found myself having silent conversations with Mill as I worked through his autobiography. We talked deep under the clear blue Caribbean water and high up on those dry rocky hills of Virgin Gorda.
It hit me on one of those hot August walks that Mill was what was missing from the treatment. The drama had no foil or buddy character. Who would be a better buddy than urbane and hyper-intellectual John Stuart Mill? The dialogue was easy to imagine since so much of it was right there in the Autobiography and in On Liberty and Utilitarianism and in his many other writings. It was a pleasure to reread them.
Nor should it be a problem to encode the gist of Mill’s unique prose style in a sufficiently powerful adaptive neural fuzzy rule-based system of the future.
We already had the algorithms. The computing power would surely arrive someday thanks to Moore’s Law doubling of computer chip circuit density every two years or so. I went on to publish the precursor of just such an adaptive rule-based system in a 1998 issue of the MIT virtual-reality journal Presence. Neither the editor nor the reviewers seemed to mind that the first figure in that paper shows an image of Mill contemplating the relative value of 12 famous paintings. The figure caption reads in part: “Here the evolving utility surface forms in the ‘mind’s eye’ of a neural fuzzy agent based on nineteenth-century English philosopher John Stuart Mill.”
I also owed Mill for a research result that I published in 1986 called differential Hebbian learning. This is the trick of estimating causal links in an AI causal network or updating synapses in a brain by correlating the changes in the nodes or neurons. Earlier Hebbian learning schemes simply correlated the activity of the nodes or neurons. That leads to a lot of spurious causality. It also leads to networks overgrown with useless synapses. Correlating changes or velocities gives far fewer cases of spurious causality. It also gives a crude form of an arrow of time. The first page of that 1986 paper credits the differential-Hebbian idea to Mill’s method of “concomitant variation” for inferring causality from observations. Mill published that method of causal inference in his 1843 treatise titled A System of Logic. As he put it in Mill-speak: “Whatever phenomenon varies in any manner whenever another phenomenon varies in some particular manner, is either a cause or an effect of that phenomenon, or is connected with it through some fact of causation.” I just put over-dots on two symbols to denote time change. That differential-Hebbian technique is now the standard way to grow fuzzy cognitive maps from data. It is also the basis for the modern theory of the differential synapse in computational neuroscience.
So what better way to repay the favor than to make Mill one of the dramatis personae?
The novel came together after a few more dives and walks with Mill. There was only the matter of writing it and then rewriting it.
No one knows what the next world war will look like. The dark clouds of nuclear and cyber-warfare proliferation continue to grow. A bad omen was North Korea’s successful test of a hydrogen bomb in 2017. A more subtle omen was the first retail sale of a recreational drone. The steady march of AI will only make bombs and drones more lethal and harder to detect and shoot down. That same march of AI will only make it ever cheaper to attack than defend.
The next world war may look nothing like the one in Nanotime. I suspect that it will look far worse.
Bart Kosko
Los Angeles
April: 2019
Prologue
The year is 2030. There are 10 billion people on earth. In 1990 there were a trillion barrels of oil left in the ground and the world burned 24 billion barrels of it a year. Now there are fewer than 100 billion barrels left. Much of that oil will be gone in five years and will require massive deep-earth and offshore drilling. Countries in the Middle East now own about 85% of the world’s oil reserves.
Oil profits have reached new highs even as oil efficiency has reached new levels. Gas is $20 a gallon in 1990 dollars but gas cars can get 100 miles to the gallon. Most cars in Europe and Japan run on superconducting batteries. Most of the world’s electricity still comes from ever scarcer fossil fuels and uranium. And most Western governments have long since defaulted on their national debts.
The amount of information in the world has doubled every year or two since the late 1990s. One second of computing time on the linked computer nets of 2030 involves more CPU cycles than all the CPU cycles in all the computers running in the year 2000. These vast machines compute with electrons and photons and plasma but still need power to run. Oil provides most of that power.
The oil can run out but the factories and information machines must not stop. The search for alternate fuels is the gold rush of the age. And the age is the End of Oil.
Chapter 1
Dhahran
Saudi Arabia
The sun set red on the dusty oil fields of Dhahran. The young man walked from the fleet of oil tankers floating in the oily green waters of the Persian Gulf. He walked next to the fresh black asphalt road and then took a dust path up the rocky hillside and into the red Saudi sunset. Adu wore blue jeans and a black tank top and a white head wrap stained with sweat. Over his back he carried an old woven bag. Adu looked straight at the red sun and did not squint.
The dust trail led to the control station. Adu had walked this trail hundreds of times to bring his father oranges and figs and outlawed plum brandy. He liked to sit with the old man and seek his counsel on Allah and women and hear his tales of the old ways and the old wars with Iraq and Egypt.
The pump station was a large mirrored box in the forest of steel oil derricks. A lone satellite dish sat on its black-pebble roof and now the dish too shined red in the sunset. The eye could miss the mirrored box in the day. The Saudis hoped the mirrors would also fool the fine eye of a smart cruise missile.
Adu stood before the seamless plastic door. His father had told him as a boy to say “open sesame” three times to himself and the door would open. He stood still now and waited and once more the door slid open. He felt the cool air spill out and heard the wild rhythms of the synthesized Arabic rock music.
Adu walked in and nodded to the two guards behind the black security desk and heard the music change. The guards had tuned the music to match their moods. The music slowed and softened when they looked up at Adu. Then the music returned to its former volume and driving rhythm.
; The two guards nodded to him.
He walked past the guards and could see their hands fondle a flat gray panel. A small nude blond Playmate danced and screamed in the air above the panel. She was a blue-laser holograph from the new issue of Playboy. Adu smiled at their mischief as he stood at the next door and punched in the prime-number code on the cipher lock.
Adu found his father in the silent control room. The old man watched the green display panels that showed each oil drill cutting through clay and bedrock. Brown panels showed the oil volume of the capped wells and their current market value in dollars and marks and yen. The old man held his left eye in a tight squint and still wore the khaki pants that Adu had bought for him last year on his weekend vacation in Abu Dhabi.
The old man stood when he saw his son. Adu bowed slightly and gave him the woven bag. The old man smiled and reached in the bag and pulled out the largest of the three oranges. He held the large orange under his nose and rolled it in his fingers. Then he squinted both eyes closed to savor its citrus smell.
Adu spun and struck the old man with a perfect backfist.
Adu’s first two knuckles crushed the old man’s right temple and ruptured his brain. The old man died on impact with his eyes still closed. His corpse slammed against the brown display panel and bounced to the floor.
Adu turned to the computer console and tore off his white head wrap. His scalp was bald and crisscrossed with fresh thin pink scars. Adu closed his eyes and bowed his bald head to the console.
Hundreds of wireless signals filled his mind’s eye. They were the same signals that the antenna dish on the roof saw. The first signals appeared as snow or static spread over wide bands of frequencies. The snow bands passed through one another in complex schemes of multiplexing.
Adu willed his own wireless commands to access the computer and use its bank of decoders to despread and filter the signals. Then the snowstorm gave way to a parallel montage of sound and images: news broadcasts of sex scandals and plane crashes and African famines. Interactive lottery drawings and bouts of casino gambling. Nature programs of ocean poisonings and new deserts. Sports programs from soccer to group bungee jumping and killer kick-boxing. Televangelists who preached modern versions of the Semitic religions. Soft-core and hard-core porn among all the races. Old movies and videos from the 20th century. Weather and smog maps of the Middle East. Hundreds of pie charts and poll results from hundreds of millions of on-line voters in the Middle East and around the world.