and GooglePlus you’d exhaust yourself. So, one winner emerges. Data scientists
attribute the success of these winning nodes to “preferential attachment” – the idea
that if Brian Arthur is using Microsoft Word and I’m using it you are likely to do so
too. But there’s another secret: More widespread adoption makes the whole system
faster. Winner-take-all marks that network hunger for the compression of time. 234
There’s an additional feature at work in the very newest of these billion-user
clusters that’s worth our attention: It’s not merely that we’ll use them because
everyone seems to be doing so, it’s also that as more users weave themselves into
each others’ lives and the machines into too, these nodes of power get smarter.
Google Maps can predict the fastest route from your house to your office because it
can watch the movements of hundreds of millions of users, each silently pinging
their location and speed as they creep through rush hour or sprint down an empty
motorway. As more people use GPS enabled devices, the quality of this data gets
232 The network magnetism: Chamath Palihapitiya, “How we put Facebook on the
path to 1 billion users.” Udemy Growth Hacking: An Introduction lecture published
January 9, 2013 and available on YouTube.
233 Winner-take-all: van der Hofstad, p 24
234 But there’s another secret: Albert-Lazlo Barabási “Network Science”,
Philosophical Transactions of the Royal Society A: (2013) 371
161
ever better, like a video resolving itself from low-quality to HD in front of your eyes.
Success attracts still more users. All of them are Google’s sensors, in a way. Medical
diagnosis, cybersecurity, trading algorithms, search – pretty much any linked ball of
chips and humans and sensors throbs with this logic. The best of the leading
technology firms understand the power of this logic: Google’s TensorFlow artificial
intelligence engine, for instance, was largely regarded by experts as nearly a decde
ahead of competitors in 2015. So the company began giving away access for free. In
traditional economic terms this would be insane; but with network logic the
strategy is clear: The more people who use TensorFlow, the smarter it gets, which in
turn attracts still more users. Dense and learning fusions of mind and data like
TensorFlow and other soon-arriving AI systems are all gated universes. The
“increasing returns” for those inside – you, me, our neighbors – breed mutual
efficient success and, of course, massive power for their owners. We’re part of the
game too: The more people tied in, the better our lives get.
The topological charm of these explosively growing clusters was first teased apart
by the electrical engineer Bob Metcalfe in the 1970s. Metcalfe was hunting for a
better way to send data – say grocery lists to his wife – through Menlo Park and he
perfected a connection protocol called Ethernet, which soon became a standard for
linking machines. What Metcalfe noticed, as more and more users piled into the
gateland of Stanford’s Ethernet-connected machines, was that the power of the
system was growing exponentially with each additional user. This became known
later as “Metcalfe’s Law”: The power of a network grows, massively, with each
additional user. A system with one phone, for example, is really not very useful. Who
would you call? A system with two phones means one possible connection – we can
call each other. But when you increase the number of phones by a factor of two –
from five to ten, say – the number of possible connections more than doubles from
ten to 45 . The difference between Bob Metcalfe and his wife sharing grocery lists
and a connected national network of husbands and wives is immense – an insight
that led Bob Metcalfe and his wife to start a networking company that made them
billionaires.
Metcalfe’s Law has another angle, and it’s here where some of the unnerving,
dangerous political power of network gates is revealed: It’s not merely that the
power of a network grows exponentially with each additional user; it’s that the cost
of being cut out grows every bit as fast. Maybe even faster. If I shut you out of Google
today, it’s painful. But tomorrow – after a day of new information and websites and
services come on line – it will be even more costly. The network scientists Rahul
Tongia and Ernest Wilson have called this “The Flip Side of Metcalfe’s Law.” 235 To be
excluded from a database of cancer genetics, for instance, when it has a million
members is probably not such a painful problem; to be locked out of the chance to
compare your genes with a billion others, however, is hugely costly. Maybe even
fatal. Imagine if I cut you out tomorrow from the NYSE, your phone system, smart
diagnostic webs, cybersecurity patches? It’s not like you can swap our your Ford for
235 The network scientists: Rahul Tongia and Ernest J. Wilson III, “The Flipside of
Metcalfe’s Law”, International Journal of Communication 5 (2011), 665–681
162
a Dodge. There is nowhere to go. Gatelands produce “winner take all” systems, but
they also produce these “loser gets nothing” dynamics – and an absolutely chilly,
fatal cost of separation from the winning system. A discipline of network science
known as “queing theory” helps us understand why. In studies of massively
connected systems, the more time machines spend on their main task – hunting
prime numbers or DNA patterns, for instance – and the less time chattering with
each other about how they will compute, the faster they run. Winning protocols
avoid this terrible inefficiency of translation because it can be spread across so
much connection. In fact, the great breakthrough of computer systems in recent
years has been the ability to handle massive amounts of data all at once, to maintain
versions of information in a concurrent state many places in the world. This is the
essential technical leap that permits compression of time. And it depends entirely on
careful and gated design.
To be inside a gated system is, then, really to be faster because of the slickness of
communication that becomes possible. The very structure of the system accelerates
that compression of time. This design feature of networks, in which winners take
more and more, is why gates, and their careful use, will become the most dramatic
lever for business, research or international politics in coming years. It also explains
why our modern gates are different than older ones. Why it is so damn costly to
leave them; why mastery of them is even more insanely lucrative than Cecil Rhodes’
gold mines. Think of the old industrial age power games for a moment: Britain and
Germany tried to match each other with their industrial output during their fatal
competitive sprint 150 years ago; but imagine if network effects obtained? If
Britain’s initial head start in the industrial revolution had given them 90% global
trade share? Germany would never have even tried to compete. They would have
been the Friendster of the 20 th Century: Isolated, slow-growing, powerless and
finally consumed by the winning system.
5.
Networks crave gates. Once a billion people were connected together, of course
there was going to be something like Facebook, a Gateland where they could link
one to another in ever-thicker cascades of connection. Once everyone could record
and watch and share videos, something like YouTube was inevitable. As we try to
picture the world ahead of us, as we try to ask what tools of power we can acquire to
twist this dangerous landscape into something we can manage and predict and
control, we must ask of it: What does it want? The world wants a protocol for the
fast exchange of money. It wants a basic language protocol. It wants a place to swap
information about IT security holes. It wants instant translation systems to replace
the need to learn to English or Chinese or Spanish so the world can move yet faster. I
believe it wants certain sorts of alliances, a particular type of superpower and even
craves a new form of politics. For any nation that controls these gates, there is a
possibility to use that position to create still more gatekept platforms, to shape the
protocols that tie platforms together in the way roads or jet planes link the physical
world. This is the iron law of Gateland: Connectivity is power. Which means that
gatekeeping is, at the end of the day, our most powerful point of a control.
163
It won’t surprise you that, in recent years, for instance, the world has seen an
acceleration in the construction of physical barriers, of fences and walls running
between nations and defining and in and out. Roy Hassner and Jason Wittenberg,
two American political scientists, scored out the pace of global wall building and
found a sharp acceleration: In fact, of 51 national enclosures built since the end of
World War Two – the Berlin Wall being the most famous example – more than half
were constructed in a rush of self-protection between 2000 and 2014 236 . And more
are coming: Hungary, Kenya, Algeria and India now posthole their borders in initial
exploration of what might be built. There’s a frantic urgency to some of this. The
Spanish government, for instance, raised a ten-foot high, razor and camera topped
fence around their Saharan footholds in 1998 – the enclosed land was controlled by
Madrid, so it was technically “Europe,” which made an irresistible target for wouldbe
migrants. The fence wasn’t enough to stop the flows. So they built a second one to
run around the first in 2001. Then, in 2005, thousands of desperate Africans
launched a coordinated charge against. A couple of dozen migrants died in the
attempt; a thousand made it through. The Spanish responded with a third line of
fence, this one 20 feet high, electic, camera-watched.
This pattern of ever more stacked defense is repeated everywhere. The walls, fences
and trenches of the modern world seem to be getting longer, more ambitions, and
better defended with each passing year, Hassner and Wittenberg concluded. Unlike
traditional lines of defense, the Maginot Line or the Great Wall of China for instance,
the aim of 21 st century barriers in places like Israel or the US or Spanish Morocco
have been less to stop a rolling armor blitzkreig than to slow the insidious
movement of smugglers and spies and criminals, or the hopeful dashes of fleeing
refugees. There’s an affective and – to those on the inside – appealing asymmetry to
these borders. They are mostly marked and built by richer, more modern, more
stable nations desperate to control in and out. The creation of gates is, we should
sense now, the corollary of connection. That they can sometimes be piled into the
“winner take all” efficiencies Brian Arthur first teased out should give us a sense that
there is a logic to this emergence.
Reviewing the problems of deadly disease contagion after the 2015 Ebola pandemic,
Bill Gates examined this connection-and-gate lemma in the sharpest, most
worrisome historical terms. “There is a significant chance that an epidemic of a
substantially more infectious disease will occur sometime in the next 20 years,” he
wrote. “In fact, of all the things that could kill more than 10 million people around
the world, the most likely is an epidemic stemming from either natural causes or
bioterrorism.” 237 This was the cost of a fast-moving, interconnected world. It was
what floated free from the extension of Paul Virilio’s line: Airplanes produced the
236 In fact: Roy Hassner and Jason Wittenberg, “Barriers to Entry: Who Builds
Fortified Boundaries and Why?” International Security, Vol. 40, No. 1 (Summer
2015), pp. 157–190
237 Reviewing the problem: Bill Gates, “The Next Epidemic — Lessons from Ebola”
New England Journal of Medicine 2015, April 9; 372:1381-1384
164
airplane accident. Well, networks will produce the network accident. Many, in fact.
A world primed for contagtion, Gates suggested, needs – and is missing – more and
better gates. (Yes, here you can pause to double take at the serendipitously strange
fact that the wealthiest citizen of our gated age is named Gates, just as the richest
man in the era of pulling oil out of rocks was a “Rockefeller”) Connection demands
systems to gaze ceaselessly for the smallest sign of assault or need for change or
accidental shimmering movement of danger. Part of what made the Ebola response
successful is that the response really was in the form of gates as we’ve come to
understand them here, not walls. Protocols for biological reaction, for medical care,
for epidemic monitoring, the urgent helicoperting in of support and aid and ideas.
The gates assembled around the Ebola pandemic were its solution. Had it merely
been walled off it would have spiked, grown, mutated and finally escaped.”The
world,” Gates wrote, “needs to build a warning and response system for outbreaks.”
This is true for all the outbreaks we face – whether they are disease or financial
panic or terror. There will be, you can be sure, ignorant calls to build walls and not
gates. This misses the point of networks. In an age of interconnection walls are
nearly as dangerous as no walls. What matters is bulding topologies designed for
gating, for milking power from the profound logic of networks we’ve seen at work.
For decades after the 1929 financial crisis triggered an historic global depression,
economists and politicians debated what had gone wrong. What had they missed?
The world, it emerged, had been wired with an economic system designed rapid
movement, but politicians and bankers had forgotten to put in the needed brakes.
They had tried to run an industrial engine against the background of gasping
political structures. Basic adjustment mechanisms – release valves for financial or
currency pressures – had not been invented, refined or installed. 238 When we find
ourselves in coming years gasping through some sort of financial crisis of our own,
or running scared from a cascading military or social epidemic that slaps us
unexpectedly, what do you think the most likely cause will be? It will be a failure of
gates. Too few in some places; too damn many in others.
Remember the haunting, persistent vulnerability of our age. A twitch anywhere in
the system can rock and even crack the whole edifice. Space Time Compression’s
ability to shrink distance to zero and time to an instant can turn anywhere, anytime
into a battlefield or a spot for revolutionary innovation. Any moment can present us
with questions of war and peace. Today we have no central theory of gatekeeping,
no ideas about balancing inside and out. We fly on instincts, but they are the old
industrial ones. Anyone can see a system that is full of holes and inconsistencies. Can
you see a way through it, though? Pressed by political, technical and economic
desire – to say nothing of the basic urge for self-preservation – more gates will be
with us from now on.
238 Basic adjustment: Ben S. Bernanke, “Money, Gold, and the Great Depression”, H.
Parker Willis Lecture in Economic Policy, Washington and Lee University, Lexington,
Virginia March 2, 2004
165
“Computer systems are, like many other things in engineering, constructed by
composition,” computer researchers F.X. Lindner and Sandro Gaycken have written.
That line, intended for switched-on digital systems, in fact fits most of our world.
Nearly everything around us, from cities to telephone networks to refugee waves,
are assembled by composition. Curriences are layered with encryption. Refugee
dreams carved from photosharing and refined with GPS. Composition in this way
breeds new vulnerabilities, new points of contact and this rather astonishing result,
one we’re not prepared for in any sense: “In a composite system there is no critical
gate,” Lindner and Gaycken explain. “Everything is a gate.” 239
239 “Everything is a gate”: Lidner and Gaycken, p. 56
166
Chapter Ten:
Defense in Depth
In which gates, operated with our new instinct, become at once a tool of prosperity and
survival.
If you walk for a few minutes to the south of Tinanmen Square in Beijing, leaving the
tomb of Mao Zedong behind you, there’s a small lane that runs back into a warren of
anonymous, white-walled buildings. The streets are unusually clean; the
surveillance cameras unusually – even for Beijing – dense. The neighborhood is
home to many of the last of the generation of Communist Party cadres who joined
and supported its rise decades ago. Several years ago, I found myself settled into one
of them for an afternoon tea with Huang Hua, one of the great modern Chinese
foreign policy figures – and a member of that early revolutionary generation. Huang
was, in a sense, heir to the Warring States diplomat Su Qin, whom Master Nan had
encouraged me to study. Huang had penetrated the madness of Mao’s revolutionary
era to see the possibility of a different order, one he’d brought to vivid life after
Deng Xiaoping ascended to the Chinese leadership in 1976. Huang had been the
country’s Foreign Minister and later a Vice Premier. He was always calm, with an
easy and relaxed temperament. One of my favorite images of him was from the mid-
1970s when, while sitting on a flight to the US from Paris to take China’s seat at the
United Nations, he was ambushed by Walter Cronkite. Huang is completely
unflustered in the scratchy video of that encounter. He sits quietly in a cloud of
smoke. Cronkite pesters him. Huang smiles, offers a cigarette to the news crew, and
though he is in the midst of a transit from the poverty, chaos and smashed politics of
China he is nothing but serene, a statesman – not the nervous representative of a
twitchy power.
“Do you know the difference between Western and Chinese thinking?” Huang asked
that afternoon as we sat inside his courtyard house. Leaves were turning outside. He
was in his early 90s then, still tack-sharp. “You see, when Chinese want to do
something, we begin with the question: What is the nature of the age? Westerners
begin with the goal. What do they aim to achieve?” Deng’s foreign policy, one Huang
shaped and executed, had been an excellent example. Mao, who ruled China from
1949 until 1976, had a darting, paranoid, murderously strategic temperament. The
nature of his age, he was convinced in his Marx-addled mind, was one of zhanzheng
yu geming – war and revolution. From this first principle, everything followed: He
honeycombed Beijing with bomb-proof tunnels, relocated Chinese industry to
isolated and gaspingly poor mountain strongholds, reacted to foreign ideas or
influence with a snapping electricity – and was bent on protective isolation as he
dragged the country through one impossible and failed isolationist development
initiative after another.
Deng, when he came to power in 1977, read the nature of his age too. He read it
differently. “There is no possibility of a great war. Don’t be afraid of it, there is no
risk of it,” he assured a group of Chinese cadres during a chat in 1983. The cadres
were having a hard time replacing their Maoist paranoia with confidence that China
167
might safely open, develop, and change. They mixed the crouching nervousness of a
secret political party with the vivid nightmares of a nation that had been invaded,
abused – humiliated, they said – by nine countries since the mid-19 th Century Opium
Wars. “We used to worry about war and talk about its possibility every year,” Deng
told them. “It seems the worry was overdone.” The world was, he felt, entering an
era of heping yu fazhan – peace and development. Terrible, nation-demolishing wars
would not soon victimize China, Deng thought. Nation building was the nature of the
age. Its tools would be science, finance, and trade. If the Chinese people worked
hard, he promised his incredulous listeners, they might by the year 2000 grow their
thin $250 per capita income to the nearly unimaginable target of $1000. “I don’t
care if the cat is white or black,” he famously observed, “as long as it catches mice.”
Socialism? Capitalism? No matter so long as it produced progress. Deng’s judgment
proved out. There were no major wars. Development was, for China, the name of the
age. The cat caught $1000 dollars of per capita income nearly on the old man’s
schedule. It is still playing with the string.
There is something admirably direct about the Western way of problem solving, of
course. The shortest distance between two points is a straight line. It has its uses. But
it’s also true that this sort of elementary geometry does not describe, always, the
best path. Particularly not today, in an era when topological collapse or instant
connection means that the distance from a to b can go to zero (or infinity) in a single
sudden shift. Yes, you may be stringing a line from a to b, but if you’re putting it up
in a tornado? Is the best route to a new Middle East really the tipping over of stable
states? Is the quickest path to climate control voluntary compacts? Are faster
markets and more unwatched capitalism the safest way to ameliorate our
unbalanced economics? Sometimes a more indirect line is called for. Sometimes it is
faster, safer. Huang Hua’s question – “Where is the world going, after all?” – turns
out to be the first essential one for any endeavor. If you are not looking from the
highest possible altitude, Chinese will warn you, you are not seeing the full picture.�
The nature of our age has, I think, emerged clearly, vividly and broadly for us now.
Constant, instant, everywhere connectivity taps with the efficient mercilessness of a
steel hammer on the glass of most of our comfortable institutions. The Seventh
Sense has let us understand a bit about the origins of this creative violence. And it
has shown us too that our era’s dominant feature is not merely cracking disruption,
but also the construction of new orders. Da po, Da li the Chinese might say – great
destruction and great construction. This is the nature of our age. Tap, tap, crack. And
then: Gatelands – speed-bred by technology and filled with possibility. That some of
these gated worlds contain billions of people or trillions of sensors, computer chips
or network nodes only strengthens their inexorable grip. That they can be so easily
or even accidentally weaponized marks the edginess of our problem. That they will
be built atop the sharp-edged shards of a collapsing older system heightens the
urgency at which they must emerge. To be inside, we’ve seen, delivers the lightspeed
benefits of time compression, of life on a nearly alive mesh. To be outside? It
is to be cut off.
168
I said earlier that I wanted to lay the Seventh Sense onto the problems of war and
peace and grand strategy – and while the insights our new instinct reveals are lifegiving
for business and eye-opening for culture, it’s to the devious and unavoidable
problem of global order I want to turn now. In a now-famous 1986 speech “You and
Your Research”, the Bell Labs scientist Richard Hamming set three questions for
anyone embarked on the exploration of new ideas: “1. What are the most important
problems in your field? 2. Are you working on one of them? 3. Why not?” 240 Well, the
most important problem in the field of global affairs is the question of the future of
world order, of how it will emerge – and what sort of aspect it may yet present. It is
the problem to be working on. It also touches, like it or not, whatever the most
important problem in your field is: Opening new markets, educating your kids,
planning five years of corporate growth. Remember the distinction from earlier in
the book, the difference between living at an ordinary moment and an historic one?
The difference between Warsaw in 1539 and 1939? In one age, history is irrelevant;
in the other it nearly throttles you. You can’t avoid being touched. We live in an
historic era, not least because the connections that define our age means epochal
quakes in one part of the system will rattle other bits too.
A feeling for history – Nietzche’s old “Sixth Sense” – should fire up invisibly in us bit
now as we think about what life in an historic age might mean for us. The surprises
in our news every day. The creaking of our old institutions. What do they augur? The
political scientists Charles S. Gochman and Zeev Maoz once made a survey of history
to see if they could spot just when the world slipped from calm moments to
epochally unettled ones. They surveyed nearly 1,000 wars, big and small, hunting
for patterns as you or I might look for trends in stock prices or sports scores.
Increasing complexity, they discovered, almost always shook the world into conflict.
States, armies, political groups, newly independent nations, ideological and religious
forces emerge and then collide in their young energy, each driven by different aims
and values and dreams. A cycle marks the process. First, disputes over small matters
accelerate – as they are doing now. Everything seems open to question: Do you
control this trade route? Who says we have to respect this arms limit? The
combined pressure of so many simultaneous disagreements announces a shift in the
pattern of all relations. A logical loop follows: “The frequency of disputes,” Gochman
and Moaz explain, “appears to be an early indicator of system transformation…The
rise in the number of disputes seems to indicate a decline in the degree of consensus
on the ‘rules’ of the international order. This heightened dissensus (often
culminating in large-scale wars) is followed by restructuring.” 241 Pressure on the
rules. Small cracks. Large scale war. Re-ordering. Like water moving from ice to
liquid to steam and back, this cycle has a predictable logic. It ran in the age of
Bismarck and culminated in World War One. It spun again in the age between last
century’s great wars too, tripping into a second World War. And it ran as well in the
240 “What are”: Richard Hamming, “You and Your Research”, Simula Research
Laboratory (Berlin: Springer-Verlag, 2010) Chapter 6
241 The politial scientists: Charles S. Gochman and Zeev Maoz, “Militarized
Interstate Disputes, 1816-1976: Procedures, Patterns, and Insights”, The Journal of
Conflict Resolution, Vol. 28, No. 4 (Dec., 1984), pp. 585-616
169
fracturing of ancient China during the Spring and Autumn era or in the Roman
Empire in the age of the Caesars. Our current shifting – that tap, tap on every
nation’s politics, media, and economics – isn’t really so unpredictable after all. 242
You may know the old saw of military parade evaluation: That the more magnificent
a nation’s uniforms are the weaker it’s army usually is. But the gold-braided
admirals of some three-ship navy reflect a very human need for security, and
particularly for a self-decided feeling of security. Every nation has it’s own foreign
affairs aims. Each cherishes a certain national image, memories of military glory and
of “interests” inseparable from culture or identity. The goals of the French and the
Turks, for instance, each evoke an encyclopedia of history, tradition and politics.
Uneasiness in Paris about capitalism and immigration, for instance; or Anrka’s
worry about ethnic division, fundamentalisms, and the creeping nuclear progress of
their neighbors. Our era’s revolutionary logic will shape choices in every nation
differently. But I’d like here to discuss American foreign policy. America plays a
central role in world order now. The country’s leading position makes her, to some
degree, an unavoidable gatekeeper. In the Napoleonic era, nearly every revolution
or war could be tied to energies emanating from Paris. During the Cold War most
puzzles of politics or geography might be framed in terms of a zero-sum competition
with the USSR. In our own age, we’ll find most every problem links to networks and
their new logic. And – for now at least – to America. Thomas Paine’s memorable 250
year-old assessment, that “The cause of America is the cause of all mankind,”
touches this linked universalism in a way he never could have imagined.
“Betweeness” or “Centrality” are the way network science labels and measures such
a role. Just how essential is a certain nation or trading platform or point? The
networks of America won’t, in the future, be the only networks. They will be less
“between”. But they will be, always and indelibly, the first reference for design.
If the traditional aim of American foreign policy was to prevent the emergence of a
challenger that threatened the country directly, or that might fence off and
manipulate Asia or Europe or Latin America against Washington’s aims, the concern
now is different: Mastery of topological destiny. There are those who observe, as an
American think tank noted in 2015, “Today the United States faces no existential
threat.” 243 This is wrong. The emergence and shape of networks is just such a
danger. Security, the great foreign policy theorist Andre Wolfers once wrote, is “the
absence of threats to acquired values.” 244 The networks around us, as we’ve seen,
are demolishing older values. Often because we want them to. Any revolutionary
technology unbuttons an older order this way. But we now approach this
destabilized world with a sort of wideyed panic. Control terrorism, manage climate
242 Our current shifting: See also Lars-Erik Cederman, T. Camber Warren, and
Didier Sornette, “Testing Clausewitz: Nationalism, Mass Mobilization, and the
Severity of War” International Organization 65, Fall 2011, pp 605–38
243 “America faces no existential threat”: See James Dobbins, et al. Choices for
America in a Turbulent World (Santa Monica: RAND Corporation, 2015) xiv
244 In his classic essay: Andre Wolfers, “’National Security’ as an ambigious
symbol’”, Political Science Quarterly, 67 (1952) 485
170
change, lock in great-state peace. So many desires scatter our minds and when we
fail to achieve them they erode our will. Here is a basic lesson of accomplishment in
any human domain: Better for us to focus on a single aim, and tighten our efforts
around that one aim and use success to strengthen our will. The answer to the great,
newly-charged existential question of politics – “Are you a gatekeeper or the
gatekept?” – is not yet decided for the United States. So, strategy first: We live in a
connected age where power is sliding into that gated order we’ve learned about.
America’s role? Not to make the entire world democratic. Not to try to bomb or
force stability onto every nation now collapsing under network pressures. Not to
passively wait for our version of history to “win”. Not to thwart the rise of other
countries. No. The strategy that fits the nature of our age and position is to perfect
and then operate the essential, topological mechanisms of power now. Sometimes
with warm humanity; other times with the ruthlessness demanded by the touch of
danger and evil. This involves gates. It involves firmness and nerve and willpower. It
demands what we might call Hard Gatekeeping.
2.
Everything around us now is or soon will be connected. Access to our trade, the use
of technology and networks, our defense or our currency – all of these must be
considered as gated systems. Each will come to throb with new power as a result of
their links, as a byproduct of their escalating sensitivity, intelligence and power, and
from that in-or-out distinction. Much remains to be instantly, fully linked in this
way. Our education. Our medicine. Our military. Real-time information, sensors and
machine intelligence will erect Gatelands in each of these areas and in others. These
walls will go up even faster than real-world berms and fences. The strategic position
of any nation or terror group or business is not going to be secured by industrial
measures. No one at Google wants to build a newspaper. No one in Al-Qaeda is
trying to float an aircraft carrier. Rather, it will be determined in the construction of
new tools, which largely means the building of a new, gated order.
“Every new age and every new epoch in the coexistence of peoples,” the historian
Carl Schmitt once wrote, “is founded on new spatial divisions, new enclosures, and
new spatial orders of the earth.” Schmitt was contemplating history: The great
Mediterranean empires and the long, millennial power trails of Asian chieftains. But
his insight touches every age, our own included. To make a nation, to build an
empire, is to draw lines, to push at borders, to smash at gates. Some of this is the
wild martial ambition of a Napoleon or Hitler (who Schmitt unwisely advised); often
it is simple protection. Lord Balfour, serving as Foreign Secretary, contemplated this
problem from his imperial vantage in 1918 and observed: “Every time I come to a
discussion – an intervals of say, five years – I find there is a new sphere we have got
to guard,” he wrote. “Those gateways are getting further and further away from
India, I don’t know how far west they are going to be brought by the General
Staff.” 245 This problem of where exactly to put the gates confronts us now as well.
245 “Those gateways”: Quoted in Robert Jervis, “Cooperation Under the Security
Dilemma”, World Politics Vol 30, Issue 2 (Jan, 1978) p. 169
171
Gatekeeping, as we’ve seen, isn’t simply the act of marking geographic lines. Those
hurrying walls and berms and triple-fences of the real world are a kind of
gatekeeping but they are also, I’m pretty sure, expressions of a deeper desire. Each
is a real-world defense against topological pressures of population and ideology and
violence, all shaped by network magnetism. The likely futility of such gates is clear
enough when you contemplate the forces they are intended to keep out: Refugees
see prosperity miles on their phones. Little wonder they rush at it. Internettransmitted
messages light up ideologies, protests or fundamentalisms that pull
people back and forth endlessly across physical borders. The world responds, as
anyone without a Seventh Sense might, with walls. But to focus merely on the
physical? This is to misread the issue. Gatekeeping includes real-world borders, of
course, but the statesmen of the future will act particularly on topologies. They will
work not merely to stop up flows, but to decide their movement. Gates will, for
instance, serve as bespoke disease buffers, created to track and then freeze the
worrisome pandemics now crawling at us on the connections of our age. Other gates
will be economic tools: The will help isolate and then soothe the unstable and
apparently broken, middle-class eroding financial patterns of the past decades.
Recall how the design of computer or network systems affects the real world, like a
marble slipped under a carpet? Gatekeeping’s topological design is like placing
marbles (or boulders) in particular ways, to direct movements and flows as you
might guide water with a canal. The great historian Arnold Toynbee once recalled a
passing moment with British Prime Minister Lloyd George during the endless (and
disastrous) Paris Peace Conference of 1919. “Lloyd George, to my delight, had
forgotten my presence,” Toynbee wrote, “and began to think aloud.
‘Mesopotamia…yes…oil…irrigation…we must have Mesopotamia; Palestine…yes…
the Holy Land… Zionism… we must have Palestine…; Syria… h’m… what is there in
Syria? Let the French have that.” 246 This sort of charmless arrogance – The Holy
Land….We must have that – doesn’t much suit our age, but that comprehensive view?
What oil and irrigation and Suez were to Lloyd George, emergence and data flows
and gates are to us. There’s a crucial difference, however: We need not take the most
precious territory of our age through invasion or colonial exploitation or in misbalanced
peace settlements. We can build the essential landscape of power. We can
build it for defense. And then we can grow it, perfect it, and attract others. Schmitt
was right, human history is the story of enclosures – and nothing about our
emerging age suggests any other approach can, with a single clear line, solve the
vexing network x network x weapons problems we face. Hard Gatekeeping means
molding the landscape, gate by gate, to encourage certain movements and make
others difficult, costly, maybe even impossible. Any future order must begin with
this first principle: The line between in and out is as essential as what goes on
inside.
246 Lloyd George: Margaret MacMillan, Paris 1919: Six Months that Changed the
World, (New York: Random House, 2002) p. 381
172
The goals of Hard Gatekeeping are simple enough to state: To protect those inside
the gated order, to make security and innovation more efficient, to accelerate
certain kinds of connection and dampen others, to manage vulnerable links to the
non-gatekept world and – perhaps most important– to use that “in or out” leverage
to relentlessly affect the interests and plans of others. These aims are the root of real
security in any age. “In anarchy, security is the highest end,” Kenneth Waltz wrote.
“Only if survival is assured can states seek other goals such as tranquility, profit and
power.” 247 Survival – and Waltz’s other essential aims – will be decided by the
nature of our future connection. In this sense, Hard Gatekeeping produces a decisive
change in our posture. It lets us aim for something instead of merely being against
movement or – hopelessly – against the future. It establishes priorities, clarifies our
real core interests, helps us budget our effort and our expenses – better, perhaps, to
spend a trillon dollars building tools of topological mastery like infrastructure or
new technology than chasing Middle East peace. 248 Gatekeeping does not transcend
the long-standing debates between the cold realism of Stalin and the idealism of
Woodrow Wilson so much as it sets us a preliminary task: Shaping the environment.
To build and control the next generation of gatekept systems should be our aim
now. In the future paper currency wil be replaced by secure digital bit transactions.
Genetic information will demand smart, live-connected platforms for efficient
mining and study. Cyber and biological security will each come to be defined by
high-speed, machine-intelligent protocols. All of these are gatekept cores. Mastering
them will be as important as having military bases, trade missions or treaty
arrangements once was. The development of such systems plays to our strengths;
they deliver both security and useful pressures of innovation. Remember – the aim
isn’t to build walls. It is to use the logic of networks and gates, and all the complex
productive force they release.
The elements of Hard Gatekeeping can be stated simply enough:
3.
First, Does it work for us? We seek a landscape of technology, trade, finance,
knowledge – all the tools of power – that balances the pressures of politics with a
colder grasp of longer-term national interest. The protocols, alliances, treaties and
data webs that enable markets, trade or thinking machines – all are gatelands we
can cultivate, improve and protect. Our gatelands must be designed to limit the
rippling risks of contagion or error moving wildly, like spring weather, inside linked
complex systems now. Control of topologies is what air superiority or sea mastery
once were. The opening attacks of future wars will come invisibly and silently
through networks, in space, in fast topological strikes – not with noisy land
invasions or bombing runs. Well-built gates will give us time and leverage. They will
247 “In anarchy”: Kenneth Waltz, Theory of International Politics, (Reading, MA:
Addison-Wesley, 1979), 126
248 It establishes: For a discussion about why a grand strategy is of practical use see
Hal Brands, “The Promise and Pitfalls of Grand Strategy,” Strategic Studies Institute,
External Associates Monograph Program, 2012
173
deliver the honestly-earned confidence of the secure. They will provide purchase for
the patient work of diplomacy, a marked shift from the crisis-to-crisis lurching of
today.
The texture of our values should be everywhere woven into these gatelands we will
cultivate: Democratic choice, freedom of thought, privacy. We won’t have time for
leisurely deliberation about ideals in the crashing, time-compacted crises ahead, so
we should begin with them. Let other nations measure gated enclosures for their
different values, marked by their paranoias and historical burdens; our strength will
be an order that reflects our habits of civilization. The curious, open-minded
searching native to our temperament, for instance. This will force constant
innovation and evolution, an antidote to the failure to change that poisons most
closed systems. A continued absence of gates will undo us in the short run. But the
long run will be a disaster if our gates are inflexible, too closed, incapable of upgrade
or – worse – girded by promises and fears that are not naturally our own.
A second principle is that we ought not force anyone else to use our gated systems.
Gatekeeper or gatekept? No more profound, painful, liberating or enslaving political
choice now exists. Nations must be free to select, in as much as they can, their own
terms of enmeshment. Our aim should simply be to build the best order we can.
America’s tremendous economic and technical lead serves us – for now, at least – as
the isolating stretch of the Atlantic and Pacific once did. There’s no need to force
others to follow us. Recall Greshams’ famous “economic law”: The way in which the
bad drives out the good. A stock market made of swindlers won’t attract investors
for very long. Well, in many networks, a sort of reverse-Gersham’s law applied. The
good drives out the bad. By this I don’t mean merely the self-cleaning properties of
systems like Wikipedia. But of linked systems more broadly. If one nation runs a
DNA database where results are skewed by rules of political power – family
members of leaders get special access for instance – it will make it less effective, and
therefore less appealing as a gateland. So we should be relaxed about letting Europe
or Russia or China try their own gated systems. Their desires reflect sensible,
understandable urges. If you were leading a nation would you rely on Washington’s
trade or financial commitments? Because of the nature of power now many nations
may have no other choice; but forcing them to enmeshment could bring shattering,
unnecessary pressure. Let them divide themselves from us; it will only weaken
them.
We need not evangelize, invade, or compel our way to power. Consider the case of
global trade. For decades after World War Two, many nations sought independence
from economic entanglement. They yearned for autarky. They scorned trade. And
they nearly bankrupted themselves in the process: Want to buy Portuguese cars?
French computers? Indian bikinis? I didn’t think so. The whole point of trade was to
solve such imbalances of skill and trade. When the age of connection began after the
Cold War, most countries made a decisive shift: They arranged their economies to
be attractive to global financial flows. Trade grew twice as fast as the world
economy. The cost of being left out was, and is, a nearly impossible burden.
American systems, if they work for America, should have the same magnetic appeal,
174
the power-law dynamics Brian Arthur found in the most successful network
arrangements. Getting it right is enough of a task. If we can manage that, what we
build will run with unimpeachable legitimacy. Chosen, not forced.
A third principle is that our gates and protocols and webs ought to be open to
others, but with each new link weighed, added to the scale of strategic balance with
exacting care. An overarching, systemic view should inform us. Nations that want to
use our cyber security systems, for instance, should join research cooperatives – and
possibly shoulder as well some problems of global network health like the
implementation of IPv6 or DNSSEC, or any of the future jointly secur standars that
will emerge. They should even – here is where the leap comes – be required to
cooperate on nuclear proliferation, cybercrime, or trade norms. The puzzle here is
clear: How to use our strengths to address our dangers? While we can’t directly
press our lead in cancer research to stop national flirtations with nuclear bombs or
cyber crime, it’s probably true that a cost of weapons proliferation should be
exclusion from life-giving technical and trade and other gated orders. Remember
that feature of network systems now, how the cost of exclusion grows even faster
than the benefits of inclusion? If the aim of any society in this connected age is to
accelerate the compression of time for its citizens, than life outside the best gates for
this will be a nearly fatal political cost. Do you really want to rely on French
computers? Wear Indian bikinis? Do Iranians want to rely on Iranian cyber-security
tools alone? It is ever-easier for citizens of nations sliced free from vital networks to
feel what they don’t have.
The courage to leave some nations out of the order we’re constructing – and to cut
others out – doesn’t come easily to us. Today we don’t really engage in a
comprehensive, linked strategy. Iranian proliferation and Chinese finance and Saudi
military support are, for all practical purposes, separated in our contemplation of
the world. Hard Gatekeeping frames these together as a piece; it spots the
topological string running between each and gives us a firm position from which to
pull the threads we want. We will come to know, to pay, the costs of over-inclusion.
But this also means that we must be concerned with the security and safety of
others, on our network or on others. To attempt to achieve absolute American
security, to do so at the expense of other nations – this will not only erode the
credibility of any system we build. It will also make them fundamentally less
effective. We are, after all, what we are conneted to.
Fourth, no nation should be permitted to force another inside of its own gates. India
might develop a great search engine; they should not be allowed to force
Bangladeshis to use it. China may develop clever trade channels to Europe; they
should not drive their use with military pressure. This principle is the flip side of our
own commitment not to compel others into American gatekept systems. It
represents an important nod to sovereign power: Each nation decides on its own. If
we’re intent on avoiding some of the chaos of the last major system change, which
produced the Thirty Years War, we should learn its lessons. That fight
fundamentally touched on one question: Who decides what happens inside a
country? The specific problem of Europe 350 years ago was if each King could
175
decide the religion of his state. The treaties that settled this debate in 1648 – after
the most devastating wars in European history at that time – established Cuius regio,
eius religio as the governing principle. Whose realm, whose religion. The very notion
of a “state” is itself being eroded as topologies, migrations, and superfast data links
eat at old borders. But nations remain an essential container of power. A rapid
collapse of that system would be a disaster, and is frankly unlikely. To be American.
French. Chinese. This still matters.
Do we demand that other nations use our protocols? Do we force China or Germany
to rules of transparency, data tracking of their citizens and research that we obey?
No. Cuius reticulum, eius reticulum. Whose realm, whose network. But, at the same
time, should we permit China or Germany to force another nation into its trading
regime? It’s biological or cyber networks? To the question, “What would America
fight against in a gatekept world?” one answer is this: We would resist any attempt
to force-fit a nation to a gated order. We should prepare for difficult, expensive
fights to maintain this principle. Russia watching neighboring states seduced into a
Chinese trading order or an American technological system won’t bubble with silent
acquiescence. Arrangements that tilt trade to one side of the Pacific will unnerve
China or Japan – or the US. But, in the end, the control logic of cuius reticulum buffers
us, as a sort of law, against a wilder madness of collapsed nations.
Fifth, we should not permit the emergence of any means to destroy our system. The
very efficiency of connected architectures makes them vulnerable. Networks, by
design, have holes that particular modes of attack can exploit. Contagion. Strikes
against central nodes. 249 Arms racing of a certain type. Our first attempt to limit risk
should be defensive. Better gates. But we’d be foolish to stop there. Gates, the
Trojans would remind us, are not enough. When truly existential dangers emerge –
nuclear weapons, certain types of artificial intelligence or genetic engineering – then
we must attack. Hard Gatekeeping should include a coiled, prepared readiness to
strike with the same light-speed movement that defines the dangers we fear. Of
course we should try to deter such attacks. But we should be realistic. There’s much
we can’t deter. And the cost of attack is awfully cheap. We won’t be truly safe
without battle plans and diplomatic gambits that can hammer at the topological and
real world vulnerabilities of our enemies. After all, they will be hammering on ours.
Our defense task is simple to say, hard to achieve: From the largest symmetrical
superpower in history, America needs to become the greatest asymmetrical one –
capable of devastating, instant action anywhere in the real or topological world.
Network scientists who study the power of “centrality” have shown the way that
certain positions in a topology – think of a rainforest, a travel network or a stock
exchange – wield unusual influence. Centrality of a certain node in linked systems
can determine control, it can produce “super spreaders” which instantly shape
249 Strikes against: I.A. Kovacs, A-L. Barabasi, “Destruction Perfected,” Nature
(News & Views) 2015, 524, 38-39
176
linked topologies 250 . Contagions, panics, economic brain freezes – these are the
essential tactics of our future. The maintenance of our centrality on the topologies
we rely on is, then, the essential part of our future security. Interrupting this is, of
course, exactly the edge our enemies seek against us: To infect us – with panic,
terror or some virtual or real virus. To alter the landscape with one sharp razor
slice, not a hail of bombs.
There is no excuse for us to not prepare to struggle in these new terms. Our military
must be shifted for the demands of such instant, cross-space attack. But it’s more
than that. In recent years America has gotten into a habit of reaching for our military
tool, and while we’ve strengthened it to an historically unprecedented degree, we’ve
let other skills atrophy. Our real aim should be to stop our opponents from adapting,
evolving and connecting to us. If we suffer some strategic, nation-crippling defeat in
the future, I don’t think it will be because our military was weak. Rather it will be
because we’ve not mastered all the other tools of topological control. The first sign
of our real awarness of this will be when the old, foolishly vulnerable systems and
tactics and ideas are replaced. We should never build another aircraft carrier
without a sense of new, networked naval plan. The idea of spending hundreds of
billions for manned bomber planes is absurd. We should embrace self-imposed
limits, we should force ourselves to innovate both with our military and in the
creation of new tools of policy. Doing so will prevent any number of conflicts. And it
will stiffen our temperament to unconventional thinking.
There’s a lean efficiency to these five principles that define Hard Gatekeeping. The
strategy echoes, in its clarity, in its clean frugality, the postures of some of the most
enduring orders in human history: The “defense in depth” of the Roman Empire, for
instance. The protective order of Tokugawa Japan. The walls of Tang China, part of a
chained national history that celebrates heroes as “defenders of the nation,” not
attackers. Gatekeeping resists unnecessary profligacy. It limits the need to
proselytize or colonize or force others to our way of thinking. And it has a universal
aspect: Hard Gatekeeping can – with a little imagination and some discipline – be
laid on puzzles like terrorism, cyberattacks, US-China relations, order in the Middle
East, the proliferation of nuclear weapons, income inequality and global warming. It
provides a sharp single image of what we’re after – and guidelines for getting there.
It’s clear enough what we need to do, I think. Build gates. Use them. Hold inside
them and protect the curious, innovative and revolutionary spirit of our people and
our age.
Let’s use this approach to examine some particular problems, and let’s start with
US-China ties – a puzzle whose resolution may be the decisive act of policy in the
4.
250 Super spreaders: Manlio De Domenico, Albert Sole ́-Ribalta, Elisa Omodei,
Sergio Gomez & Alex Arenas, “Ranking in Interconnected Multilayer Networks
Reveals Versatile Nodes”, Nature Communications April 23, 2015
177
next century. How might Huang Hua, with his nature-of-the-age consideration of
strategy have regarded the matter? What does gatekeeping tell us?
The dominant view of future relations between Americans and Chinese generally
runs along a nervously familiar historical track: An established power (the US) and a
rising one (China) consider each other. Do they cooperate? Scratch at each other in
constant annoyance? Each nation possesses a different image the world. Each holds,
as well, distinct pictures of their role. One power has enjoyed a long period of
prosperity, has built and defended a global structure; the other, trod mercilessly
underfoot by history, feels the unbearable flow of power’s rising sap, a hunger for
recognition and release and for some, revenge; her economy craves resources, trade
routes, and markets. So Germany encountered and then attacked Great Britain at the
turn of the 20 th Century. Japan collided, similarly, with Russia in 1904. France taking
on Austria, Russia and Great Britain in the Napoleonic age.
Even if the leaders of the United States and China intend to avoid conflict, they face
one of the most sobering, if not depressing, problems of international politics: The
“security dilemma”. The world is a dangerous place. So countries do things to feel
safer. Their populations demand it. Germany looks at Britain’s imperial navy in the
1890s: London could snap Berlin’s trade arteries. So the Kaiser orders two
battleships. Britain builds three in response. Germany turns to submarines. And so
on. Each country, chasing her own security, ends up less secure. This is the
“dilemma”. The puzzle is like one of those woven wicker finger traps: The harder
each side pulls to get out, the more stuck they become. America in 2012 pivots her
military gaze to Asia. China feels encircled. She orders a couple of aircraft carriers,
paves remote islands into military bases. America flies aircraft closer to Chinese
waters. Are the two sides more secure? Collisions, arms racing, accelerating distrust
become the threnody of contacts. Each side seeks something impossible: Perfect
security. Move first, certain forces begin to whisper in each capital.
The essence of this problem today, as the United States and China consider their
future, is not merely or even mainly in the details of trade, territory, or cyber
disputes we read about. These hot points are important, of course, but they are
symptoms of differing judgments of the world, of a fracture at that first step of
asking, “What is the nature of the age?” Americans generally believe the existing
global power map is just, sustainable, in need – perhaps – of minor adjustments.
America engineered this system; benefits tremendously from it’s fictionless
operation. China regards that same layout differently. As broken, unbalanced,
rocking towards a seizure – and, anyhow, built, lubricated and run without her
participation. This view is sharpened by ideology, national psychology and the bald
fact that no global mechanism ever functions perfectly. The larger nature of the age
is, many in Beijing feel, Da po, Da li. Great destruction. Great construction. Certainly
this is true inside China; it must be true outside. The nation has it’s hammers out.
They see the paradoxes of power we read about earlier, the inability to act and the
collapse of American credibility, as clearly as people in Washington or Damascus or
Moscow. Surely the global system should be adjusted, they think. Tap. Tap.
178
So though no inherent enmity exists between the two countries, a crack emerges.
Each side finds reasons not to cooperate. A shared interest in peace – neither side
wants a collision, both understand the costs – is shimmied apart by a perception
gap. Cultural, historical, temperamental and ideological differences accelerate the
unzipping of even the best intentions. Misunderstanding bends and then snaps
irrevocably into mistrust. So yes, the recovered islands and the trade tensions and
the political theories have to be managed because each is a blister that limits
forward motion – but the real source of friction? It is the scraping of different world
views. This is deep strategic misalignment. It is, in fact, the stuff of war.
In recent years, America has had two different, nearly opposite approaches to China,
each however marked by an assured national confidence that the current world
order is sustainable. One policy is short-handed as “engagement” – or, colloquially,
“panda hugging.” As China becomes more prosperous, this approach suggests, she
will become more congenial to American interests. More cuddly. This was, for some,
the basis for bringing China into the World Trade Organization in the 1990s. Later it
justified deep commercial links and educational ties. The logic of such an approach
fit the mindset of many baby-boomer Americans: China’s attraction to MTV,
McDonalds and Mercedes hinted at a still deeper desire they thought. To be
Western. China’s citizens would surely become more attracted and committed to the
world system that had delivered all this prosperity. The country would become
more global, in a sense, and less Chinese.
In fact, as China has grown in recent years she has become more global and more
Chinese. Growth, wealth, an admixture of confidence into the old national habit of
insecurity – all of these have encouraged a searching exploration. The demands of
domestic politics, the instincts of the Chinese Communist Party and the inevitable
practical questions of security and reform all produce sensible questions. China
needs oil, for instance. Must she project power into the Middle East? Control the sea
lanes of the Malacca Straits? There’s no natural, inherent resistance to American
interests in any one problem, but that hopeful American logic of “engagement”
ignores much about China’s essence. The country presses at rules she mistrusts (as
America once did). She struggles with the inevitable challenges of a political system
still smoking a bit from recent revolutions – and, of course, facing the pressures of a
networked age. But she holds, strongly, a sense of Chineseness. Can we honestly
expect to use mere prosperity to change China in a predictable, pro-American way?
“Engagement” feels to Chinese like a demand to become more American. That
prospect is as appealing as a request to become more Chinese would be for
Americans.
An alterative strategy to “engagment” is known as “containment”. (While not
formally called “Panda Kicking”, this is often what this policy looks like.) This view
takes it for granted that as China becomes more powerful she will become a threat.
Any growth in Chinese power means a decline in America’s influence and security. A
stronger China in Asia, for instance, means a weaker America. Countries like Japan
or Korea or the Philippines will be required to choose a side at some point, at some
awful 21 st Century Melian moment. The aim of containment is to squelch China
179
before such a fork appears. This approach, too, has its logical problems. Boxing the
largest economy in the world hardly seems sensible – and is likely not achievable.
Anyhow, early attempts to do so have shown how little leverage the US has. When
China announced plans to build an Asian Investment and Infrastructure Bank in
2014, the US opposed it and then watched as nearly all of Washington’s traditional
allies raced to join.
At the same time, China’s policy towards the US has been hard to decipher for many:
Is the country intent on amity or enmity? It’s clear enough that the American
options presented to Beijing are unappetizing: Change or be contained. Messages by
China in response have tickled the natural historical pessimism of American political
leaders. China speaks of wanting a new kind of relations with America, of cultivating
rich technological and financial ties, but the country at the same moment challenges
easy assumptions about how foreign investment, educational or security
cooperation might mesh together like gears of friendly progress. It’s hard to do
business in the Mainland. The military grows with astonishing speed. Domestic
habits of smash-mouth competition rattle at more sedate Western habits. Any
element of China’s actions are explainable and even sensible when seen with
sympathetic eyes – but the same exact facts can be stitched into a story that suggests
constant damage to American interests. China’s position is, to be honest, a difficult
one: The country is opaque to most foreigners, wrapped in habits and instincts that
run along such different lines that even simple daily decisions – how to run a
meeting, for example – can be a source of disagreement. In the west you prize
diamonds for their clarity, Chinese will say, in China we prize jade because it is cloudy.
There’s a charm in opacity; even a bit of sensible political magic at times. But China
is moving out into the world very rapidly, now: Fast and dimly understood don’t
fuse into easy trust.
In one sense, then, it’s possible to sense the tense, angry energy of an impending
collision. Beijing and Washington are enacting the old historical pattern. Wrestling
for years over small issues, never quite coming to grips, slowly pulling at each other
until a final, fatal collapse. Germany and Great Britain, all over again. But: Connection
changes the nature of an object. So much in our world looks different after connected
dynamics are applied. Might connection change the nature of enmity? We are, after
all, part of the same connected skein. We’ve concluded that America’s greatest
threat is not China or Russia or terrorism – but the evolution of the network itself.
That same topological evolution menaces Beijing too. If China is not our biggest
enemy, if we are not theirs – what might be developed? Might the rat trap of history
remain unsprung?
An American “China Policy” shouldn’t be regarded as some distinct initiative; rather
it must be a part of a grand strategic program of Hard Gatekeeping. Our goal aim
isn’t some shimmeringly perfect “China Policy” so much as it is to cultivate a grand
strategic disposition that informs what we undertake in cyber, Russia, the Middle
East, and China. Imagine we had a terrific “China policy” and a backfiring “Nuclear
weapons policy”? Masterful grand strategy is marked by a comprehensive elegance.
Metternich’s old construction, “A Concert of Nations” tells the story pretty well: A
180
symphonic effort, each note and instrument in place. Grand strategy tells us what to
do at every moment.
Any policy, then, must begin with a summary of our aims: “We believe the world is
entering a period of profound change that demands a shift in global arrangements.
We are building a gatekept order consistent with our values. We welcome others to
participate in it – but with conditions. We will resist attempts to force nations into
any gatekept order; we will fight any forces that threaten disruption of the overall
system.” The China policy that emerges from this view is clear enough: We will not
contain China. We will not force China to change. Rather we will develop a profound
gatekept network for our own operation, one that layers together all the concerns of
economics, trade, security and technology – as if we were designing a technical
system for fast, constant links. We will include other nations that want to be a part
of it, as long as they are willing to bear certain costs. China is welcome to develop
her own system and see whom she might attract inside. She will not be permitted to
compel nations to join. And: If China wants the benefits of a US-led order, then
Beijing must be prepared to support the maintenance of that order. That means
deeper cooperation on everything from nuclear proliferation to the establishment of
new international bodies for connected finance, biological research or nuclear deproliferation.
Such an approach resolves, in an instant, the contradictions of our current policy.
Here is our gated order, we can say, join if you wish and on the terms we both consider
best. We should remember: China has much to worry about. Disorder would land on
her own hopes for peace and development with a profound, possibly crushing
pressure. The international order is not working as well as it might. It does not
appear engineered for new stresses. It is in cooperation for the reform of that
system that the two nations share an urgent interest. And it is in a successful
defensive posture of a gatekept, network order that Washington can cultivate the
linked, leveraged moves that will encourage this to happen. While there might be a
temptation to play this game in that old Lloyd George sort of way – “We need the
Middle East. Grab it. We need digital currency mastery. Take it!” – such efforts will
only assure final, fatal collision between the two powers. We need a gatekept order.
Build it. That is enough.
What the Seventh Sense statesman sees, in the way a Seventh Sense entrepreneur or
terrorist or researcher can, is the pulsing network logic that is at work breeding a
new world – like it or not. And our policy has to aim for that future state, not for
what we have today. If this sounds difficult, you should know that in most complex
adaptive systems these kinds of adjustments frequently occur. Rainforests. Traffic
jams. The human brain. Anyplace connections are knit and reknit at light speed can
exhibit a pattern of rapid redevelopment: The world changes. The network and its
various pieces hum, in response, to a new order. Climate shift in the Amazon, for
instance, produces drier air. The pollination cycle shortens. Flowers must adjust by
growing longer leaves to protect their moisture. Birds that rely on the flowers
develop longer, sharper beaks in response. In this way bird and flower each become
better suited to a warmer age. They co-evolve. We see this sort of interaction in
181
many worlds. Data companies face increasing bandwith demands. They develop
new ways to transmit video. Application developers create fresh tools to compress
their signals. The whole system gets faster. Such co-evolution is common when
linked systems press into one another. 251
If you look back at biological history, at our planet’s seething life at it’s most
profound moments of transition, you find co-evolution in nearly every case. The
biologists John Maynard Smith and Eors Szathmary, in their masterwork The Major
Transitions in Evolution, chronicle the miraculous march of life towards ever-greater
complexity – from cells to humans to societies – as a story of ceaseless, successful
co-evolution. “Complexity is hard to define or to measure,” they write. “But there is
surely some sense in which elephants and oak trees are more complex than bacteria,
and bacteria than the first replicating molecules.” The essential, facilitating feature
of our biological hop-scotch from mitochondrial snot to Beethoven sonatas has been
an ability to change together, not merely to compete. While we humans tend to
consider evolution in murderous you win/I lose Darwinian terms, the biological
world around us seethes with a more nuanced logic: Cooperation is every bit as
important to survival.
Smith and Szathmary identified eight major transitions in their work – the evolution
from single cell to multiple cell organisms, for instance. And in each case they find a
kind of cooperative, co-evolving logic serving a catalyst to survival. “The
applicability of the concept,” Smith and Szathmary conclude, “extends across the
whole of social and natural sciences.” Inside cells. In ecosystem webs of coral or
rainforest. In traffic management. Successful evolution was, they found, always coevolution.
“It should need little persuasion to acknowledge that links we make with
others change in time,” they wrote. “That all of us age, that our roles in life evolve,
and that the society we are a part of may itself be subject to transformations on a
global scale. Coevolutionary rules aim to integrate these processes.” The world
changes. I change. You change.
Network age systems, in their effortless bridging of divisions – geographical,
technical, commercial – bounce with a particularly powerful fusion. “It is often
assumed that biological agents are by nature rationally selfish, and that therefore
external circumstances somehow have to compel them to cooperate,” the historians
Nathale Mezza-Garcia, Tom Froese and Nelson Fernadez have written about the
organization of societies over the past millennia. That sense we often have, that
we’re one Darwinian instant from some sort of lethal, teeth-baring competition,
doesn’t in fact explain all of human history. 253 Politics is a sort of complex adaptive
251 Such co-evolution: Matjaz Perca, Attila Szolnokib, “Coevolutionary games—A
mini review”, Biosystems 99 (2010) 109-125
253 “It is often assumed”: Nathalie Mezza-Garcia, Tom Froese and Nelson
Fernández, “Reflections on the Complexity of Ancient Social Heterarchies: Toward
New Models of Social Self-Organization in Pre-Hispanic Colombia” in Journal of
Sociocybernetics 12 (2014), pp. 3 - 17
182
system; it fuses cooperation with brutal competition as it aims for a kind of
equilibrium. After all, if we were constantly tearing into one another, we’d be
extinct, not evolved. Why did cities form? Nations? The key is a moment when
shared, possibly devastating risk of collective failure becomes apparent – those
moments when the whole system threatens to collapse right on top of all the
participants. The greater the shared danger, Mezza-Garcia and her colleagues
explain, the better the chance for real cooperation. It’s a forumula that fits what we
face now: “An increased risk of collective failure facilitates large-scale cooperation,
especially when the large scale system is composed of smaller, nested groups,” They
explain. ““Complex systems theory reveals an alternative to constant conflict.”
This is our world now: Linked clusters of markets, nations, machines – all exposed
to a risk of shared and instant collective failure. Such a structure upends an axiom of
politics that has run for centuries: Man is purely Darwinian and that survival is
determined by constant competition alone. In a network, survival is determined by
sociability, by cooperation. Linked systems drive people to agree on rules in order to
participate, which accelerates change and co-evolution. Smart cancer databases,
linked traffic systems, video platforms like YouTube – each are Gatelands that press
users together on a single platform, one that can be instantly updated and
constantly studied for adjustment. “Rich get richer” arrangements are an ideal place
for co-evolution, shaped as they are by easy linkage, exchange and connection to
outside events. As more people follow the logic inside the gates, the system coevolves.
It becomes still more fit. It’s this loop that makes Hard Gatekeeping so well
suited to an age of connection. We’re not merely putting up walls; gatelands are like
markets or public squares. They are loci for cooperation. In diplomatic history it’s
not hard to calls for cooperation. “We must get along or this war will devastate us
all!” But they don't work. Mostly you find hand-wringing about how terrible war
would be followed by, well, war. The logic of networks offers an escape from this sad
habit.
Co-evolving relations between the US and China can begin at the most obvious of
starting points: Both sides need to change. New pressures are already tearing at
each. Both nations need a new gatekept system. Our aim isn’t to “balance” China into
some sort of frozen checkmate. It’s not to tip her unwisely into chaotic and
impoverished domestic order – we are, like it or not, all connected. It is to co-evolve
together. And China? She won’t survive if she does not connect to a system that is
congenial to her needs. Fortunately, the networks are infinitely flexible in their
design. A Chinese model and an American model pose no problem of
interoperability. This is the best co-evolutionary strategy for each side. The dangers
we confront now are everywhere. The habitual American temptation is to break all
these risks into pieces, to find out where the problem is, and to flatten each of them:
“Let’s hit the terrorists, then the Chinese navy, then the Russians, then the drug
lords.” This will kill us if we keep it up. In a complex system, piecemeal attack
simply shocks the system to still greater complexity. The world, rather wonderfully,
is too flexible for brute force applied at one place. A dynamic defense, is a precursor
183
for survival, for evolution. And it would best be done cooperatively with the nation
that will, in a short time, have the largest economy in the world.
China is not really a threat to the US now. The country can’t assemble the ability to
demolish or attack America in any significant way, and it is years from being able to
sustain a military effort. It is not in China’s interest to provoke a match. They would
lose. And honestly they have endless, troubling problems to handle, many new in
human history. How do you care for 800 million aging people? Though China will
pass the US in total national income in coming years, the country may never pass
America in per-capita income. With an aging, 1.4 billion person population the
challenges of distribution nearly guarantee this. Time acts against both the United
States and China today. Each depends on global arrangements that are themselves
cracking. Each nation needs to adjust her structures for a network world. But these
pressures operate severely against China. She can disrupt and challenge and slow
the transition to a gated order if she wishes, but only at the cost of fatal distraction.
An American Hard Gatekept system will thrive with or without China. China is not so
fortunate. Her system, without America or other elements of global connection,
cannot continue on the path of reform. Great network construction – this is the
nature of our age. There is no fighting the nature of the age, as Huang Hua would
remind us. We should of course expect China to build her own systems. We can also
expect that she will try to avoid some of the costs that joining an American system
might impose. That’s all fine. Our aim should remain constant, calm and direct: Build
a gatekept system for our own use; enforce the rules we believe secure that system.
And remain clear in our thinking: The threat to the US is the evolution of the
network. The threat to China? The evolution of the network.
If an arrangement based on these undeniable facts seems unlikely, it is worth
remembering that in the twenty years before the restablishment of simple relations
between Beijing and Washington in 1971, the two sides had fought a war in Korea,
built a politics based on condemning one another, and drifted so far apart as to
share nearly no econoic connection. Bringing the countries together demanded
heroic diplomacy on both sides, to be sure. To read the transcript of Henry Kissinger
and Zhou Enlai’s first, secret meeting on July 9, 1971 is to regard a masterclass in
careful chessboard moves 255 . But the game itself demanded a clear view of the
board, of shared interest in the face of a changing world. On his very first day in
office, Nixon compiled a list of his diplomatic aims: “Chinese Communists: Short
range – no change. Long range – we do not want 800,000,000 living in angry
isolation. We want contact.” 256 The Seventh Sense returns us to this fundamental
question of balance, but in a new frame. It alters how we look at the range of options
for US-China relations – and American security itself. There are certain problems
both sides face, perhaps the most essential, that are simply not addressable using
255 To read the transcript: Foreign Relations of the United States, 1969-1976,
Volume XVII, doc 139
256 Nixon complied a list: Foreign Relations of the United States, 1969-1976, Volume
XVII, doc 3, footnote 3
184
traditional thinking. Co-evolution exists as a real possibility for the two nations but
so too does co-extinction, a problem of mis-designed complex systems that can
rattle themselves apart. This is precisely what a confident Hard Gatekeeping
strategy can avoid. When we speak of the possibiltity of a new type of great power
relations it is because so much of future power will be decided on a topological
landscapes, on that nearly living surface that is marked by such different rules than
the older, industrial power maps. It takes a network to fight a network, as we’ve
said.
Networks confront us, constantly now, with the unexpected. No one forming the IMF
in 1949 considered digital currency. No one developing arms protocols in the 1990s
thought about cyber weapons. Artificially intelligent weapons, migrant waves,
income fractures – these and other emerging puzzles were never contemplated in
our existing international arrangements. Collectively attacking these challenges is
not merely more efficient; it is our only option. Yes, it’s easy enough to picture the
US and China battling each other over islands and protocols and technology
structures in the decades to come. Certainly this possibility has to be prepared for.
But this is to regard the world in industrial terms. Ask yourself: What, really is the
point of power? It is to secure stability, not to tip the world to chaos just because we
think we might be better off on the other side. The networks tell us what they want.
They want gates. Our only question – and it is the same question that lingers in
Beijing and Washington – is: Are we smart enough to listen?
5.
How will power distribute itself in the future? Just what will the network of nations
and datawebs and insurgents finally look like when it settles into some predictable
if still roiling order ten or a hundred years hence? Will it be made up of different
walled and gated systems, linked at certain moments, unplugged at others? Or might
it collect with that “winner take all” logic we’ve seen firing along the most efficient
networks? Most traditional foreign policy looks at history and sees the constant,
violent rocking of a “balance of power”. Nations thirst for security, they vie with one
another for influence, resources, and control. France rises. Statesmen in Berlin,
Vienna and London conspire to knock her down. This is the order that dominated
Europe for much of the last 500 years. But history is also filled with enduring, stable
arrangements – moments where the balance settles a bit and a single power
dominates. Asia, Europe, the Middle East, South America – each produced nations
that stretched mastery the system for generations, for centuries even. If you map the
rise and fall of nations and empires over the past 4,000 years – as the political
scientists Stuart Kauffman, Richard Little and William Wohlforth once attempted in
an heroic statistical effort – about half of human history has been marked by this
sort of hegemonic stability. China led East Asia’s order from 1300 to 1900, for
instance. The Japanese and Koreans and Vietnamese all found it was wiser to play by
China’s rules than to attack her. Assyrian imperial arrangements overmastered a
dozen smaller states from the 9 th to 7 th centuries BCE. The Dehli Sultinate managed
hegemony in South Asia from 12 th to 14 th centuries. The Mughals owned nearly two
185
hundred years of dominance starting in the 16 th century. The Romans managed
centuries of Mediterranean control .
The secret of those long-running orders was something that will be familiar now:
Each possessed tools of power which permitted assembly of empire at an unusually
low cost in lives and gold and effort. Kauffman and his colleagues, as they
considered the results of their survey, noticed each long-lived empire pioneered an
administrative design that embodied an efficiency much like that of our own
network dynamics. The addition of new territories brought more that they cost to
masters of long imperial orders. Like new users on a social network, or Baran’s fishnets,
they married easy expansion and high returns. “Rome rose because it
combined the strengths of traditional Republican institutions with innovations that
gave it a unique capacity for inclusion of foreigners,” they explain. “Magadha was the
most administratively durable of the ancient Indian states; and Qin, with the selfstrengthening
reforms of Shang Yang – economic reforms and military conscription
as well as bureaucratic innovations – developed the most penetrating and brutally
effective state structure in its international system.” The Incas, the Han, and nearly
every long-standing empires glistened with this attractive logic. The secret to
hegemony, to avoiding a violent power shifts every few decades, is a structure that
grows without additional, destructive costs. When Machiavelli coldly called Rome a
“republic for expansion”, this was what he had in mind. Enduring empires have been
engineered, like a modern network, for growth and prosperity.
It’s too early for us to know if this logic will obtain in our age. But networks evolve,
as we’ve seen, to what makes them most efficient. They crave speed and growth.
And this means they want cooperation; it’s the essential fuel for co-evolution. The
traditional view of the international system as anarchic is not wrong, but we’ve seen
how when you snap any object into a network system it begins to crave a kind of
hierarchy. Networks change power balances. National fury and rebellious twitches
and competition will, of course, be a part of the transition ahead. But as we look
back at the industrial tools that matured and spun up the world to a war in the last
century, we can see how they they were designed in a sense for direct collision.
Massive industrial armies wrestled in symmetrical power battles. Network power
hums differently. The design logic of linked systems means they function poorly
when tuned for simple brutality. It’s why the tools for our new world are so
dangerous in the hands of those who don't understand what they are capable of, and
what they demand. We should remain fixed on what might emerge as a future state,
and on avoiding the shaking dangers of the route. It is from that posture that we can
begin to consider the most essential and interesting and profitable questions. The
most profound is probably this one: We’ve seen now what it feels like to use the
Seventh Sense to contemplate the networks around us and to examine the global
system with its risks and opportunities in a new way. But what do we discover
when, as if we were looking into a mirror for the very first time, we use this
powerful new sensitivity to examine ourselves?
186
Chapter Eleven:
Citizens!
In which the Seventh Sense rescues us from an unexpected danger.
1.
I never needed much incentive to go see Pattie Maes. Belgian, usually dressed in
some black fashionable getup, she was like a human shot of espresso. You ended
every conversation wide awake, eyes open. When I first met her in the 1990s, she
was in charge of much of the work on artificial intelligence at MIT’s Media Lab –
Danny Hillis’ old home. Maes had arrived at MIT in 1993 and almost immediately
turned to the problem of making machines that might think. One day, as we were
discussing just how the strange miracle of computer thought might occur, she
introduced me to a puzzle of her field that has stayed on my mind in the years since.
It is called the “Disappearing AI Problem.”
Back in the 1990s, as the Internet was emerging into popular consciousness, Maes
and her team were tinkering with what was known as computer-aided prediction.
This was an advance on the ping-pong conversations Joseph Weizenbaum had
coerced from ELIZA in the 1960s, You: “I am bored.” ELIZA: “Why are you bored?” In
Maes’ experiments a computer would ask, for instance, what movie stars you liked.
“Robert Redford,” you’d type. And then the box would spit back some films you
might enjoy. Cool Hand Luke. And, well, you had liked that film. This seemed like
magic at the time, just the sort of data-meets-human question that showcased a
machine learning and thinking. An honestly “artificial” intelligence. Maes hoped to
design a computer that could predict what movies or music or books you or I might
enjoy. (And, of course, buy.) A recommendation engine. We all know how sputtering
our own suggestion motors can be. Think of that primitive analog exchange known
as the “First Date”: Oh, you like Radiohead? Do you know SigurRos? Pause. Hate them.
Can you really predict what albums or novels even your closest friend will enjoy?
You might offer an occasional lucky suggestion. But to confidently bridge your
knowledge of a friend’s taste and the nearly endless library of movies and songs and
books? Beyond human capacity. It seemed an ideal job for a thoughtful machine.
The traditional approach to such a problem was to devise a formula that would
mimic your friend. What are their hobbies? What areas interest them? What cheers
them up? Then you’d program a machine to jump just as deep into movies and
music and books, to break them down by plot and type of character to see what
might fit your friend’s interests. But after years building programs that tried – and
failed – to tackle the recommendation problem in this fashion, the MIT group
changed tack. Instead of teaching a machine to understand you (or Tolstoy), they
simply began compiling data about what movies and music and books people liked.
Then they looked for patterns. People were not, they discovered, all that unique.
Pretty much everyone who liked Redford in Downhill Racer loved Newman in The
Hustler. Anyone who enjoyed Kid A could be directed safely to (). Maes and her team
found themselves, as a result, less focused on the mechanics of making a machine
187
think than on devising formulas to organize, store and probe data. What had begun
as a problem of “artificial intelligence” became, in the end, a puzzle of mathematics.
The mystery of human thought, that great unknowable sea of chemicals and instinct
and experience that would have let you place your finger on just the song to open
the heart of your date, had been unlocked by data. Here was the “Disappearing AI
Problem.” A puzzle that looked like it needed computer intelligence demanded, in
the end, merely math. The AI had disappeared.
For several decades this accidental digital magic trick – “Hey, where’d the AI go?!” –
bedeviled machine intelligence. It gave the entire thinking machines enterprise a bit
of an occultish flavor. Many problems that once seemed to demand the miracle of
thought really only needed data. The human was still doing the thinking; the
computer was simply computing. It was extremely easy to draw a line between
where the human ended and the machine began. This was a puzzle that had been, in
a sense, anticipated at the very dawn of the digital revolution by the mathematician
Alan Turing in a 1950 paper called “Computing Machinery and Intelligence” that he
published in the British journal Mind. “Can machines think?” Turing began. 257 His
idea was to test this question in the following way: Have a research subject – a
secretary, a graduate student, anyone – chat with an invisible interlocutor by way of
a keyboard. Then ask: What are you connected to? Another human? A machine?
Turing figured you could call a machine “artificially intelligent” if it could fool a user
into thinking it was human. “Please write me a sonnet on the subject of the Forth
Bridge,” Turing suggested a tricky user might ask. What computer could possibly
know about this famous Scottish landmark; to say nothing of being able to rhyme
“Forth”? When the response came back, “Count me out. I could never write poetry,”
you’d think that sounded awfully human. “Add 34957 to 70764,” Turing suggested
you might fire back. Say the computer pauses. Then, thirty seconds later, “105621.”
Are you dealing with a clever machine here? A dim, honest, slow-multiplying
human? Impossible to tell. The distinction between machine and man blurred a bit.
Buried in the very premise of Turing’s test was an assumption about what machines
could do, and how they might do it. “The idea behind digital computers may be
explained by saying that these machines are intended to carry out any operations
which could be done by a human computer,” Turing said. His smart machines would
be tuned by humans to do human tasks, in a human sort of way. Write poetry. Do
math. His test of artificial intelligence was designed to figure out if a computer could
think like a human. You and I might be able to spot patterns in movie habits given
enough time, but as more complex problems emerge, as a world of a trillion
connected points becomes a sea of data to examine, there is no chance we’ll match
the machines. This opened a then nearly unimaginable possibility: What if, some
day, a computer could think better than a human. Could, in a nanosecond, come back
with that elusive rhyme for “Forth”: North. For such a machine, passing Turing’s
Test – thinking like a human – was an interesting challenge. But more interesting
257 “Can machines think”: A. M. Turing, “Computing Machinery and Intelligence”
Mind 49 (1950), 433
188
was a machine that could think in ways a human could never understand, let alone
achieve. For such a device, to pass Turing’s Test or slip past a Voight-Kampff check
machine will be trivial. In his 1950 paper, Turing sensed the possibility of this
development – and the crisis that might ensue. Could man handle the crushing
sensation that a device was outperforming him? Perhaps dramatically. “We like to
believe that Man is in some subtle way superior to the rest of creation,” Turing
wrote. “It is best if he can be shown to be necessarily superior, for then there is no
danger of him losing his commanding position.” What would have surprised Turing,
I suppose, is the speed at which we’ve acquiesced – and even accelerated – this very
loss of our dominance.
Imagine a device poking at the very origins of the universe at a speed of trillions of
calculations a second, spinning past Newton’s and Einstein’s laws and into a realm
of physics apprehensible only inside its own electronic consciousness. Compacting
time – centuries of human scientific labor reproduced in moments – before shooting
far ahead, alone to a subtle knowledge we can only envy. Such a machine would not,
as Newton had, stand on the shoulders of giants so much as it would muscle its own,
unique way ahead. The AI would have disappeared, but to a place very different
than where Maes’ AI had gone. Hers had been erased by human design. This new,
really “thinking” AI would slip to invisibility because of its own light-speed
cognition. It would think itself out of our understanding. No human could follow,
limited as we are by our wet, slow, decaying biological software. Humans and
computers, after all, deal with information differently. Think of how poor your
memory is compared with the perfect fidelity of a machine, or the way people can
even “remember” events that never happened. The machines would have more than
knowledge, then. They would linger close to a possessing a profound and
inscrutable wisdom. They would inhabit an honestly miraculous gateland that no
human would ever enter. And this is where the problems would begin.
We’ve now passed the moment when humans completely train the very best
machines. The AI devices can teach themselves, now. Of course there are still
decades of adjustment, of leaps in hardware and programming ideas to eliminate
the seams between our minds and the fused ideas of a digital system. But the
humans in the loop of the best of these designs, the hard working and well meaning
geniuses of the New Caste, are as much trainers as engineers now. They resemble
shipbuilders of an earlier era, preparing vessels for voyages to lands they will never
themselves see. Their digital minds – technically machine learning structures such
as deep neural networks – run quickly to the frontiers. Every moment, all around us,
machines are educating themselves. About the world. About themselves. About us
too. Tuned AI plays games, looks at photos, studies chemical reactions, reads your
email and watches you drive – and then it tries to unspool just what is going on,
before winding it up into a new instinct. Machine learning systems already produce
mathematical proofs that linger beyond what a human mind can understand; next
may be a machine-to-machine mathematics that expands the dimensions of thought.
(In a man and machine fusion that surely would have unnerved Joseph Weizenbaum
2.
189
of ELIZA, the theoretical mathematician Doron Zeilberger now names his computer
as a co-author of his papers. He calls it Shalosh B. Ekhad, a play on the Hebrew name
of his IBM 3B1.) 258
The AI systems designer Roger Grosse has named two paths to this sort of wired
sensibility: “Predictive Learning” and “Representational Learning”. 259 That first
approach is what Maes’s movie machine pusued. The computer is simply checking
what it encounters against a database. It teaches itself to predict based on what has
been seen before. This sort of knowledge begins with massive amounts of data and
then hunts for patterns, tests their reliability, and improves by mapping quirks and
similarities. Google engineers have a device that can gaze into a human eye and spot
signs of impending optical failure. Is the machine smarter than your
ophthalmologist? Hard to know, but let’s just say this: It has seen, studied and
compared millions of eyes to find patterns that nearly perfectly predict a diagnosis.
It can review in seconds more cases than your doctor will see in a lifetime – let alone
recall and compare at sub-millimeter accuracy. Fast, thorough predictive algorithms
make what might once have been regarded as AI disappear. The machine isn’t all
that wise; it just knows a lot.
On the other path, the one of “representational learning” the machine uses a selfsketched
image of the world, a “representation.” Computers using predictive
methods to recognize 10,000 numbers pulled from a database of scrawled hand
writing now identify 90 percent of the images. Self-trained machines, however, line
up each scanned pixel against a representation of the very idea of writing. They
screen millions of pictures with nary a mistake. Faces, disease markers, obscure
sounds – all these become scrutable not because the machines have been told what
to look for, but because they’ve sort of figured it out themselves. The AI is actually
starting to think, much as you or I might, first by building up a picture of the world
and then applying it, much as a child might build comprehension of traffic rules just
by watching Mom driving every day.
With this representation finished, these nearly alive “thinking” meshes navigate by
themselves. You can see already the competition lingering here – who can build the
most sensitive model of the world? You? A machine? Even today basic versions of
representational AIs can study a map and name the most important roads. They can
predict cracks in computer networks days before a fault. These programs take
longer to train. They are harder to program – and they demand almost unimaginable
amounts of computing power – but what emerges is a subtle, lively kind of insight. A
machine with a representational understanding of Mozart’s 41 symphonies can
write you an extremely convincing 42 nd – or, if you wish, an even earlier First
Symphony based on what it knows about his evolution as a composer. It can do it
again and again. In seconds. The basic attitude of these researchers behind this
technology runs, they confess, like this: Mozart was a fantastic composer. If he wrote
258 He calls it: See Nielsen
259 The AI systems designer: Roger Grosse, “Predictive Learning vs.
Representational Learning”, Building Intelligent Probabilistic Systems: 2013
190
even more symphonies they’d probably be great too. Unfortunately he’s dead.
Wouldn’t it be nice if we could sample his old symphonies and make new ones
whenever we want? 260
In the future we’ll invite AI into our lives to harmonize away many of the problems
we face, not merely making up for Mozart’s inconvenient mortality. “AI Agents” will
linger along side us. They will compose versions of themselves we’ll not quite grasp,
even as we appreciate their efficient magic. “AI is both freedom from programming
and freedom from understanding,” runs one programmer’s line 261 . Today machines
that once demanded millions of lines of code can function with a fraction of that.
Instructions are sent to machine learning systems; the programs do the rest. Such
designs balance their mystery with efficacy. They speak to and learn from each
other too. Part of the reason that the the “Does it think like a human?” Turing Test
will be insufficient in the future is that the machines are not learning from only from
humans. They are learning from each other.
Perhaps this is not such a bad thing. The distinguished physicists George Ellis and
Joe Silk, who spent a lifetime trying to stand on Newton and Einstein’s shoulders to
grasp answers about gravity or the future of our universe, electrified many of their
peers in 2015 with by wondering if perhaps too much of science had become
unscientific, unverifiable, unreliable. The great grand ideas of our day, notions like
string theory or dark matter, differ in a crucial way from Newton’s laws of motion or
Einstein’s principles: They cannot seem to be tested and significantly proved. And
this had fired a trend among younger physicits: Perhaps there was no need for
proof. To Ellis and Silk this seemed an awful retreat, dragging physics back to a pre-
Enlightenment age of conjecture, superstition and instinct. “This year, debates in
physics circles took a worrying turn,” they wrote. “Faced with difficulties in applying
fundamental theories to the observed Universe, some researchers called for a
change in how theoretical physics is done. They began to argue — explicitly — that
if a theory is sufficiently elegant and explanatory, it need not be tested.” Fans of such
an approach called the idea “post-empirical science.” This strange, oxymoronic idea
was, in a sense, like proposing post-rules baseball: A recipie for wild, swinging chaos
that would make scorekeeping impossible.
The strange, boiling debate did however reflect an underlying and unnerving truth:
Science does seem to have stalled. And it became inevitable to ask: Might it be
possible that the machines – or some fusion of Shalosh B. Ekhad andta human mind
– can reach into an understanding of laws that no human alone can fathom. We’ve
said again and again: Connection changes the nature of an object. Perhaps it changes
260 Wouldn’t it be nice: Andrej Karpathy, “The Unreasonable Effectiveness of