W. Brian Arthur’s lectures – Day 1, 28 Mar ’18

W. Brian Arthur’s lectures – Day 1, 28 Mar ’18


my name is Jan Vasbinder I’m the
director of para limes who organised these lectures to which I welcome you
it’s the first of four lectures by Brian Arthur and I in many ways I think
they’re historical because they talk about Brian will talk about the history
of several important developments in the past the complexity science Santa Fe and
and Silicon Valley and I forget one but the the talk about that but it’s not
only that he talks about that history he is one of the people the key people that
made that history and he is also a superb storyteller which you’ll find out
in the course of the coming four days I know Brian since I think 2005 I’m not
quite sure I’d never know precisely but it that was when we were setting up the
Institute parademons in Europe and I got a telephone call from Brian Arthur about
whom I had read in the book by by Mitchell
Waldrop who said I’m Brian Arthur I’d like to be involved in this foundation
of this institute perilous so a few days later he came to the founding session
and prior to the founding session we had a dinner and I needed somebody to give a
talk a dinner talk just to somehow give the idea what Santa Fe Institute was for
the spirit of Santa Fe was and then John Holland who was also part of that
meeting said you should ask Brian because he can go tell good story so ten
minutes before dinner I asked Brian to to give a talk and he looked a little
bit strange time to ask he said but then he said yes then he started to talk I
asked him for to speak for ten minutes he spoke for 45 minutes
it was a fantastic talk it set the spirit for the phone for the founding
meeting the next day and it’s I think it played a key role in in getting
institute polymers going in Europe and subsequently para Lemus in NTU because
there’s a direct line between Institute polymers in Europe and part of him is
here so it’s a great pleasure that Bryan has agreed to give these four lectures
but before I give the word to him I’d like to ask Steve Lansing to introduce
him in a better way than I do hi well I I’m the director of the
complexity Institute here and you’re in for a treat I mean I’m about to
embarrass Bryan at least say I think I will cuz I’m gonna actually tell this
story I’ve it’s been my great fortune to spend a lot of time in Institute’s for
Advanced Study all over the place right in Princeton
and Stanford and Durham and Amsterdam etc etc so um so and Vienna so I’ve run
into some interesting people in those places Brian as you’re about to learn in
these four lectures he’s accomplished four bodies of work any one of which
would be a major contribution to to thought in though you know in our in our
era any one of them they are extraordinary and as as as John says
he’s also a great storyteller so this is gonna be fun and it’s really I think
it’s wonderful that we will have this event because each of those is treated
as a discrete subject he’s famous for four different things and different
communities of people know about them I don’t think anybody but if the only
actually tracing the ideas that go through those things is something that
maybe we begin here as we look at the threads that connect these four bodies
at work so you’ll hear from him in a moment what those four topics are I
think let me just tell you a little bit about his background so you get some
notion of how it is possible for one man to do all these things so Brian here’s
here’s Brian’s pedigree he got his first degree a Bachelor of Science and
Engineering at Queen’s Belfast where he was at the top of the class
he got a master’s degree at the University of Michigan mathematics
followed that up with another master’s degree at Lancaster in the United
Kingdom in operations research then he took another PhD a PhD at the University
of California Berkeley in operations research and a PhD in economics but he
couldn’t have two PhDs from from Berkeley so that one turned into a
postdoc he is the first winner of the Lagrangian computer science it’s like
the Turing prize but it’s the first prize in computer science in 2008 so
he’s and he’s taught economics now wait am i guess economics is the one subject
he hasn’t taught yet he’s one of the most his perennially mentioned as a
candidate for a nobel prize in economics what he actually taught at Stanford was
demography so it’s an entry you can see this is gonna be an interesting story
he’ll start today by talking about what the Harvard Business School here
won one of the perhaps the most cited paper in the Harvard Business Review
was an article that he wrote 24 years ago on increasing the turns and the new
world of business and I won’t say anything more I’ll say one thing about
it this is basically foundational for a Silicon Valley as you will soon hear he
wrote it in with and and having written this paper as he was editing it can I
tell that story he uh he asked a friend of his to help kind of look it over so
that guy was Cormac McCarthy and you may know that name because he has a Pulitzer
for his novels he also has an Academy Award for his films he’s a very famous
writer so Carmack had a look at this thing and between them Brian and Cormac
polished it up and here it is it’s the foundation for you know things like
Google and UHN it’s an extraordinarily influential paper so we’re gonna start
today the four topics we’ll begin with let me get this right now increasing
returns economics widely considered to be the most important economic theory in
technology so lead off with that ok followed by the origins of the
Santa Fe Institute where he was the director of the first research program
at the Santa Fe Institute launched the Santa thing at the research program in
the Santa Fe Institute so that’s the beginnings of complexity science and
well here’s something about Santa Fe Institute 3rd lecture will be about
non-equilibrium economics also known as complexity economics that’s what he won
the prize for that turns economics is a field that’s based on equilibrium theory
Brian has been developing a theory that is not based on stable equilibria you’ll
hear about that ok it’s an extraordinary work and then
finally amazingly technology is leading theorists probably the leading theorists
on the evolution of technology on how technology evolves so those are the four
topics we’ll hear about them I think more or less in that order but we’ll
also be hearing about and thinking about the ways in which you know what connects
and what could possibly connect all those ideas in a work of one guy so
Brian so pleased to be here well I’m delighted to be here and I will try to
live up to what yawns I hope what Steve said just sitting there drew minds me
about my father who was a great raconteur and storyteller very very
witty and I remember him being boxed into a corner somebody said that at a
gathering bill that was my father Oh Bill’s very funny says something
funny bill I feel like this position of tell a good story Brian
anyway I’m delighted to be here I’m delighted you’re all here and I must say
I’m deeply flattered that young sets us up and flattered to see so many people I
I know and respect here fair number of faces I recognize I’d be giving four lectures the the
first is and two of them sound quite arcane I hope they won’t be because they
actually matter in the economy I’m talking about increasing returns which
is basically to do with positive feedbacks in the economy then talking
about complexity and how I could caught up in it and we might we’ll have a word
from a couple of other leading theorists and complexity in this lecture tomorrow
on Monday I will talk about a new very general way of looking at the economy
that we got going in Santa Fe and then on the evolution of technology and this
last one’s project I’ve been working on with Stefan who’s here in the audience
Stefan Turner who’s in charge of the complexity hub in Vienna so it’s a it’s
a varied menu and I’m thrilled to be here as I said and mm-hmm
let me just pitch in one of the things I want to do as I go is talk about not
just what I did or what other people working with me did I want to talk about
the atmosphere of these various different disciplines at the time a lot
of this work ranges from the early 1970s on through to today and there these this
work was done in an atmosphere its radically different from the sort of one
that’s beginning to take over these days I want to also hear their point out
things that I wish I had known roughly when I was your age I knew nothing about
science or nothing about how the whole thing works nothing about how wonderful
it could be nothing about how catty it could be
and nothing about all the obstacles you will run into an ax career
I knew basically nothing I came from a family and a background where nobody had
been at university and the only people you might know who had ever been at a
university would be the family doctor and they weren’t talking they were sort
of poking at you so I didn’t know anybody who’d been at a university that
wasn’t unusual where I grew up which was Northern Ireland Belfast there the
number of people went to university or the proportion was 4% one out of 25 or
so and I was very proud to go to University and it took me five months to
discover I was an undergraduate we didn’t use that expression we thought we
were students I’m the student at the University but in there was some
pronouncement or rather talking about being undergraduates and I thought oh my
god is that right I’d heard the expression later a bit later I decided
people teaching me we’re doctors so on so doctors somebody doctors somebody
else I thought that would be nice I looked up the rules it said you needed
at least 20 years of experience you need to be extremely distinguished you needed
to belong to the Royal Society you need to decide the other I was looking up
Diaz C’s which are a very high form of honorary doctorates
I wasn’t looking at pH DS I didn’t even know what a PhD was so please take it
that I was starting with a totally blank slate and in a state of ignorance and
not unusual in Europe at that time also so I want to talk about the ups and
downs for me the bottom line is that I have had a wonderful time fantastic time
doing research and I feel extraordinarily lucky to
have sort of parachuted into this lifetime and being able to do research
it’s I could get up in the morning go to some Jones College and Santa Fe and
think about things and then write them down and your sleuthing all the time and
you go oh my god you know that can that be true
everybody you’re probably all of you share some of that Steve was just saying
that there’s two species of hominin if I got that right that have been discovered
by people at NTU in the last short while by sleuthing and you could hear the
excitement in his voice so this is the atmosphere I come from
on the other hand so that’s the Good Fairy the day I was born the Good Fairy
says oh you’re going to have a job during research this would be wonderful
and the bad fairy shows up the bad fairy says but you’re going to have a career
in academia and he won’t believe the hassles you’re going to get into and you
won’t believe the positive things in academia for me it’s been what we call
in the UK snakes and ladders you go up one ladder and then you’re
down near the bottom I can you go up another one Americans call that chutes
and ladders they don’t like snakes and that’s been all been cleaned up so what
I do want to expound on here and there is what I wish to God I had known when I
was 24 years old or thirty years old I found out the hard way and it wasn’t
some of it was wonderful a lot of it was difficult as you’ll hear also I began to think well these are
different topics is there any theme to what I’m saying and it reminded me that
about seven or eight years ago I had I was 60 and 60 plus anyway and I thought
you know something I’ve been in engineering mathematics I’ve been in
population studies or demography I economic theory I’ve been in technology
I’ve been in complexity all those different subjects and I thought what’s
wrong with me am I just flitting around from one blossom to another surely that
means I’m in dilettante and many of you might have had the same experience or
might in the future and I encourage you to keep changing fields if you’re in
academia or in research as often as you feel like but I got quite a spawn and I
thought well I’ve wasted a lot of my time just sort of as English you should
say banging on about different topics without posing much every ten years or
every seven or eight years I changed my field and then I began to realise there
was a common theme this as much to my relief common theme was that I have
always been looking at systems that were changing about evolutionary things that
were unfolding in my fascination as time and things that unfold over time so if
you’re wondering what is he talking about this time and why is it so
different from the previous one that’s the common thread and I started to
wonder for these talks could I say there’s a common theme is there anything
that really is their message that I’ve had and I really do believe there’s a
message the message is this we and I’m talking about the 20th century but
carries over into this century we have tended to view the world as highly
orderly highly machine like all you need to do is fiddle and twist
the dials the an was trained this way to I’m sure I can see many of you were but
there’s something inside me screaming you know well the world isn’t orderly
it’s not a machine it is really an organic thing that’s
changing and one layer falls on top of another and it’s one thing after another
surely we can do better than seeing that so this is the theme that’s gonna run
throughout these talks the world isn’t perfectly orderly it’s certainly not a
machine it’s not very predictable and history matters like crazy all that
could ruled out in the 20th century in fact it was ruled out in the high
enlightenment in the 1700s why because everybody was stunned and dazzled by
Isaac Newton and rightly so the all these planetary orbits where you could
predict things down to fractions of an inch Halley’s Comet would return you
could predict all sorts of things I’ll tell you a personal story here very
quickly and added a bit of insight and this is very personal to me I don’t
admit to this very much but for but thirty years I’ve studied Taoism or Tao
and I was invited up about three years ago to Beijing by Taoist master and he
held dinner and my honor and I was very flattered and he had got about sixteen
Daoists around the table all Chinese not much English was spoken but there was a
translator and so these were think Hogwarts if you want whatever you know a
strange bunch of animals but they’re all Daoists and all Chinese I shouldn’t say
that I should and would I give a short introduction so
for about five minutes I trained as an engineer mathematician born in Ireland
and then I thought I’d connect with the audience and I said you know I was born
in Ireland but I think when I came into this lifetime I was heading for China
but somehow the wrong button got pushed and I the re-entry wasn’t very good and
I landed at the opposite end of Eurasia which was Ireland and got born there by
mistake and I everybody giggled and smiled and I thought that’s it I
connected and about half an hour later after a number of drinks he was just tug
on my sleeve and it was the master she sure foo I don’t know Sifu in Cantonese
the master grabs my sleeve he says don’t say you were born in Ireland by accident
no accident I said what do you mean you know for me it was a bit of a joke I
said what do you mean it’s very very serious about it said you weren’t born
in the West by accident you came here to teach people Tao but
Westerners can’t understand Tao so what did you do
you came into the West and you taught them Dow parading as science
he says they get science they don’t get Dow he’s so I said but science isn’t Dow
he says yes it is he says complexity science is pure Dow so I was flattered I
thought if that’s what I’m here for to show the world is organic and changing
and everything’s swirling and one things leading to another which is very much
the spirit of Taoism I thought I’m all for that
I just didn’t know who that was teaching that and as complexity and many
of us are I won’t M so there is a theme and that is that the world is basically
organic and always changing and it’s not necessarily an equilibrium it’s
certainly not optimal I want to start these talks quite arbitrarily I thought
where could I start I could start last year but I want to start quite
arbitrarily in 1969 where I was about 24 if you’re trying to calculate all this
and so start in 1969 and I’m in a helicopter and literally in a helicopter
and it’s just landing in in California and Berkeley that’s how I arrived in
Berkeley I’d been working for McKinsey and company as a management consultant
in Dusseldorf in Germany I had plenty of money and when I got to San Francisco
Liz to be a little helicopter service was terribly expensive cost $12 I forked
out the $12 to get to Berkeley by helicopter landed very grandly what I
didn’t know and wasn’t prepared for what awaited me in Berkeley give you some
ideas here talk about mind-blowing experience my
son is here so I can’t say too much about all of this but it wouldn’t
surprise him he grew up in California and I had spent two years in the u.s.
already at the University of Michigan I didn’t know John Holland was there
Kenneth Boulding was there and I think Steve Lansing may still have been there
in 1967 and back to Ireland back to Belfast the British army moved in to
occupy Northern Ireland in August 1969 we had choppers all over the place we
had armored vehicles we had tanks down the streets we had tear gas we had
the whole thing I don’t want to say I blame them Ireland or Northern Ireland
was an uproar I was so relieved when I got on the plane landed in San Francisco
took the helicopter over to Berkley and what did I find in Berkeley armored
vehicles tear gas choppers all over the place uniformed soldiers on the ground
is the National Guard Ronald Reagan who eventually signed my PhD certificate was
governor and he had sent the Blue Meanies at the Alameda County Sheriff’s
Department and the National Guard to quell all the uprising in Berkeley it
was a different cause but the atmosphere was exactly the same just before I got
to Berkeley one student had been shot in the people’s parkinson and so this was
for real I wound up in International House there are two things in Berkeley
that sort of didn’t quite jive and one of them was that along with all the
occupation by the army and by the National Guard and whatnot was where
hippies people with flowers and their hair people had come out to Haight
Ashbury for the most part they weren’t students but there is a whole hippie
culture I’d be going down Bancroft Avenue and I
remember a guys are staggering up to me and very long hair mustache is down to
here which eventually I adopted as well foo Manchu mustache and it comes up to
me so hey hey man can you tell me where the free clinic is and of course we all
knew oh just go down there turn right and then three boxes or whatever it was
when they when the authorities tried to break up the free clinic this we knew
this was serious the other thing and this is really where I want to take off
from was that there was an atmosphere in Berkeley like the rest of the United
States where if you did science you saw the world is very mathematical literally
mathematical the world was highly ordered the world everything was sort of
structured I was in operations research Berkeley was a staggeringly brilliant
place to be it had stellar people because it’s been collecting them from
the 1920s on the chancellor of Berkeley Robert Gordon’s sprawl in the 1920s
decide do you want to hire a couple of physicists Berkeley wasn’t no so he
calls up friends he said somebody at Yale or something he says who are the
best young postdoc physicists in the country and the guy says but I can tell
you two pretty good ones there’s one called
Ernest Lawrence and there’s another one that Yale his friend called Robert
Oppenheimer so Berkeley makes them an offer they can’t refuse and they come
out by car bumping along across the US and whenever it was nineteen twenty
something and start a really good Department of Physics I mean really good
to part in the physics that eventually leads to Los Alamos and the consequences
there off Berkeley was amazing I do remember meeting the artsy name and of
Neyman Pearson lemma firm I don’t know if any of your statisticians it was by
de t but they had a collection of people not to speak of Nobel Prizes that a
collection of people where you’re kind of interacting all the time not with the
top people in California but with the top people in the world not all of them
and the main thing I got out of Berkeley well among many other things by the way
I said to people that I got a real education in Berkeley oh and the
academic side was also pretty good the but what I really got out of Berkeley
was this feeling I just met whoever it was you know with
a Nobel Prize and I and the guy looked normal I mean this looked like a
standard human being well I’m a standard human being you know maybe I can do
something and previously in Belfast er Lancaster or Michigan I hadn’t quite had
that feeling so I was very very much taken with this idea but Berkeley was
extremely mathematical and they believed very much in order in predictability of
everything I was in operations research which is a mathematical subject of how
to optimize systems well once you understood the equations that ran those
systems you could tweak the equations maybe continuously over time and thereby
optimize the whole thing that’s not particularly easy I find it quite
rigorous and quite difficult my fourth year there I showed up in my advisers
office his name was Stewart Dreyfus and he was proud that by that stage he he
was on Nixon’s Richard Nixon’s enemies list because he’d been at the RAND
Corporation and he was brother of Hubert Dreyfus who had been arrested for
growing pot and was the chairman of the philosophy department and so you get the
atmosphere in Berkeley super academic but super sort of you know who you’re
sure I go into dreyfuses office he says you’ve been here four years he said
we’re cutting your funding off at the end of this academic year that’s a nine
months time I’ll give you a mathematical problem he was a mathematician he says
you can solve this problem you have HD okay I said you know so it’s like a
doctor writing prescription says here’s the problem gives me the problem the
problem was how to control things where the controls are very time delayed and
how to do that rigorously and mathematically so officially that’s
called the optimal control problem with linear dynamics the equations were
unfolding over time a quadratic criterion with Gaussian noise and time
delays think of it this way you have a satellite out near Jupiter you want to
control it over time make sure it’s pointing in the right direction but it’s
about let’s say it’s eight minutes away or 20 minutes away whatever it would be
in the solar system you do something like that it doesn’t get your signal for
10 or 20 minutes and then it responds lose oops
overdid it there it’s like trying to adjust the shower tap when there’s time
delay in the pipes some people had solved this problem
Russians mainly optimal control theory which this was a part was very much a
Russian subject and been heavily influenced and introduced by Russian
mathematician called you have pontryagin I was later to meet pontryagin
who turned out to be a rather small man and who’s blind he was a blind
topologist and his mother led him around explaining who was there and what was
but either in spite of being blind or because of being blind he had fantastic
spatial insight and had pretty much pioneered optimal control
theory in the Soviet Union it spilled over into the United States RAND
Corporation all these people who did optimal everything and landed on my
shoulders in 1992 I was given this problem a few people solved a few
special cases but the overall problem as a mathematical one was open meaning
nobody had solved it the other part by the way if the problem is that the
signals you’re getting us to where your system is maybe a satellite we’re out a
day to its signals oh I’m over here but that was 10 minutes ago so you’re trying
to figure out where it’s going where it is everything’s delayed how do you
optimize that by the way the person who’s behind that’s called root of
Kalman and I once met him too I went up to Kenneth Pharaoh’s office in Stanford
where I was working with much later 20 years later arrows there he’s got a
smile on his face O’Brian he says this is really Kalman I said oh wow one of my
heroes Kalman the math Kalman Lee the engineer cover notes of me
no I’m Kalman the mathematician so anyway this is what I was working on and
I couldn’t solve the problem I read this I read that the problem was given me in
October by Christmas I’d lost about 20 pounds
I’d love 10 kilos or so I was looking gaunt my mother came out she was worried
about my health I turned serve greenish yellow color and this was in California
I couldn’t solve the problem but I stuck to it
so Easter came I still hadn’t solved the problem and I couldn’t solve it and
finally I got my courage up and went to my advisers office and I said I can’t
solve this problem you know would you kindly write
something I’ll turn the crank and he’s this sure he says don’t be upset by the
way he had beads long hair no no no no don’t be upset he said you can’t solve
it but don’t forget nobody else has been able to solve it I sizer I took some
heart but I was really really upset and ashamed and I give it up and I started I
wrote a dissertation or at least wrote down solve some other problems where I
knew what the method ought to be and this was pretty safe about three months
later in June 1973 I was sitting in the main library in Berkeley and I was I
don’t know what was I reading something or writing down something but I’d long
given up that problem and suddenly I got this flash like you’ve got mail I
thought what I’ve you’ve got the answer to something and I said oh all right
what does that mean well you have to take certain matrix and then you have to
partition the matrix so it gives you little matrices on inside the big matrix
and then you invert the partition matrix and that’s it you have the solution what’s the problem because you know
you’ve got an answer here’s the answer I said okay but what’s the question
oh and this is all sort of poem coming into my mind you know I’m just sitting
there it wasn’t voices nobody was saying you
have it was really well oh you’ve got an answer oh all right answers a but
partition matrices getting inverted but what’s what’s what’s the problem
problems the linear quadratic control theory control problem with time
continuous time delays and I thought oh so took me another five
to ten seconds to figure out put the whole thing together the problem the
solution I remember walking out of the main library feeling very satisfied oh I
could see how to do it now for two more months they did nothing on it I don’t
know why and I think this very typical research I didn’t do anything I didn’t
try to verify it I just knew that this was correct
and I happened to be visiting my girlfriend in Washington DC mother of
someone here and I was visiting her in Washington DC and I got on a plane I was
working in Honolulu this is the advantage of California you can so I was
working in Honolulu I could on the plane I took out a yellow pad it was a 10-hour
flight from DC to Honolulu and I started to write down the problem and when I got
off the plane I had my dissertation written it took ten hours I’m not saying
that’s the way it works all the time but this was the first wake-up call I wasn’t
sure I could do research at all I hadn’t done anything I had given up the problem
but nevertheless something inside was working and I think that this is
generally true for many people are probably for all of you that things
rummage around inside you and then you know then when you have the key ideas oh
oh oh this is simple and I’m always struck by the simplicity of ideas if if
they’re correct and how obvious they are I want to turn to economics and spend
some time talking about this was also unburied in Berkeley from Germany and I
had been working for McKinsey in Germany and seeing how the German economy was
laid out we would take private planes go down the Rhine and my friend on the team
and say you know shaolin’s 51% of that the deutsche bank owns 46% of that and I
was getting aerial view of the whole German economy I came to Berkeley
wondering how economies work and I encountered my adviser his name was
David gale people he was a mathematician gale looks at my letter ISA so you’re
getting interested in the economy Gale by the way was famous mathematician
people talked about man years and gale minutes Gale had written a paper at Rand
with Lloyd Shapley called the marriage problem it was an algorithm for matching
it’s used now in college entrances all over the Gaels Shapley algorithm Gale
died before he could get a Nobel for that but Lloyd sharply his co-author
picked up the Nobel for that one so I go into Gail’s office quite austere looking
mathematician he says you’re interested in the economy what have been reading I
said Samuelson great says says Gale sounds wonderful book Samuel Paul
Samuelson they all fall terrible of economics had written this fabulous book
on the economy very arcane very mathematical showing how the economy
could be viewed in the 1940s this was his dissertation he got a Nobel
for this – you’re gonna hear a lot of these people who were awarded the week’s
Bank Memorial Prize and memory of Alfred Nobel
Samuelsson had written as treaters showing how the economics could be not
imperfectly mathematize but completely and utterly mathematize and so Galle
says great Greece has Samuelson’s foundations wonderful book and then I
said nothing and then he stared at me and the penny dropped he says you do
mean the foundations don’t you no I said I mean the elementary textbook I’d been
reading Samuelson It was as if I don’t know I’d read the comic book version so
I was off to a very bad start in economics economics at Berkeley when
this is by way of background was very much an endeavor similarly to operations
research if you knew the equations if you knew everything the whole economy
could be made mathematical one of my professors gerard debreu
yet another Nobel Prize winner believed as a mathematician and if you just knew
the axioms of the economy you could deduce everything about an economy that
turned out not to be true thank God and but to Bro was greatly honored for this
in fact David Galen said if I didn’t go to dubrow’s lectures no no no
so in Berkeley the economy was regarded as an equation system one of my
professors comes in I was studying economic development Ireland was sort of
a third world country at the time maybe a bit like Singapore in those days
developing but not very well off so I studied economic development and
economic development I’ll give you a hint of this at Berkeley and one of my
professors I won’t mention his name another distinguished economist comes in
and he says he writes on the blackboard and those
as bigs capital’s squiggle e he says let capital squiggly be the set of all
economies then let capital squiggle D of capital squiggly been said of all
developing economies and I’m sitting there I had been in places like Turkey I
I come from Ireland I’m thinking what so I walked out of that class later I
became a development economist I taught in Kuwait I taught in Syria
I went Bangladesh and Nepal in India etc etc but I can assure you they don’t work
as mathematical entities very well so how do they work
so Berkeley was very much into mathematize everything and making sure
everything was optimal one of the key things I started to look at was the
whole idea of diminishing returns and this was just stuff I was wondering at
and the lecturers in Berkeley economics was presented to me as a set of theorems
and even years afterwards I thought of economics as a bunch of theorems I think
if you’ve been luckier maybe Stefan has you you could could
have seen it more and more narrative terms than that but theorem etc and then
there was usually no proof and then the next theorem and I believe this stuff
you know anyway so the economy was regarded very very mathematically and there’s a standard view of economics and
let me just go through a very quick bit of economic history rarely or the
history of thought in economics there’s superb economist from Adam Smith and
even before that on up until about 1870 and they regarded the economy
narratively they regarded it as quite organic
was populated by human beings if you read Adam Smith or if you read later Jay
s mill or Marx who turned out to be a super economist you will see human
beings human beings and machines and what they’re doing and what their
institutions are from about 1870 onwards mathematics enters economics in
particular algebra and equations and what else enters economics is general
equilibrium theory borrowed from physics in the 1880s and 1890s by Leo Walker
whose French worked in Switzerland trained as an engineer and knew quite a
bit of physics they mathematize parts of economics and the history of the whole
20th century is the history of those parts of the economy viewed
mathematically Samuelson was a landmark in this with his foundations the book I
hadn’t read and kind of arrow and a few others arrow had been trained as a
mathematician and Kenneth arrow came in and along with Deborah mathematize and
perfected the mathematics of general equilibrium theory in the 1950s and I
was entering this about 20 years later and that’s where the field said it was a
mathematical field to make the mathematics work you had to assume in
the economy an awful lot of negative feedback think of where the economy can
be is like conflicts Bowl and the state of the economy’s ball bearing rolling
around this cornflakes bowl what you’re able to show on general equilibrium
theory if you wheel up enough conditions is that with all this concavity
of the cornflakes ball the ball will there’s only one place the ball can
finish up and that’s at the bottom of the cornflakes bowl and and that’s
optimal and it’s unique if you wheel up enough axioms and conditions and then
what if the conflicts poles traveling in a car the ball bearing gets bumped
around but in a sort of homeostasis where the ball bearings always returning
to this bottom the cornflakes bowl I don’t exaggerate I don’t and I don’t
want to make fun of this general equilibrium theory is elegant it’s
beautiful whether that’s reality
I would say well stay tuned if that’s reality but I
at the time I was a bit bothered that this was a bit too nice in particular
that showed the economy as long as there is enough diminishing returns or
negative feedback to be predictable you know it’s going to end up at the bottom
of the conflicts ball history didn’t matter ends up the bottom of the
conflicts bow no matter where ball bearing had been before it was efficient
or optimal or Pareto optimal or whatever you want that’s the bottom of the
conflicts Bowl and and all of this was super orderly but in the 1800’s
economists had started to worry what if they were increasing returns and these
were just little worries it was kind of like a sunny day and the cloud of pr2
appears on the horizon it’s okay but tiny little clouds so here’s one
economists started to think about develop economic development and what
about Italy well it’s really developed economically and if you went back to
1400 or so you’d say well places like Naples and south of Italy places south
of Florence even Rome very important and some places like Venice or Milan are
also important but you couldn’t quite say what would happen by the 19th
century northern Italy was heavily industrialised south of Italy was not
and people were starting to think well what wasn’t about northern Italy turned
out nothing was that special they started to argue the rail mechanism
people began to realize was that if you start a cluster of industries and Turin
it attracts more industry then it attracts skilled workers from the other
part of Italy that gets bled free of capital and that’s bled free of skills
and suddenly you get one part of the country highly developed and another not
you can see this all over Europe by the way but if history had been different
maybe the south of Italy area around Naples could have been staggeringly
well-off and northern workers go down to the south by the way it said that people
who have come from countries victorious in war think of England or think of
America up until the Vietnam War tend to favor that the inevitability and
optimality of history that’s where it had to be why because Britain is
superior why cos were English well as people who come from nations that were
defeated in war Ireland etc they well diamond could have been different if
history had gone the other way Scotland’s another one so economists by
the 1800s in early nineteen hundred’s were starting to wonder what about
increasing returns they realized there could be multiple solutions to the
economy economies economists call that equilibrium multiple equilibrium so
here’s Alfred Marshall the great English economist genuinely superb economist in
1891 Cambridge and he’s musing all of this is enough appendix H to his
principles of economics he says ok start n firms off all the same they’re all
identical and they all have diminishing costs meaning the more they produce the
cheaper it gets that’s increasing returns if you the
more you produce the more advantage you have because producing was getting
cheaper and cheaper and cheaper appendix H it’s not in the main text this is what
will happen he says the market will likely go to
which F this is literally the quote to whichever firm first gets off to a good
start so your cornflakes bowl now is upside
down and think of it as a bowling alley you can pull the ball down the middle
it’s an infinitely long bowling alley but the balls gonna fall off sooner or
later bumping along maybe ghusl wind will fall into this side gutter or that
one you don’t know which so Marshall says which ever firm first gets off to a
good start and then he realizes but that could be any one of them in principle
maybe everything wasn’t perfectly balanced but still was sufficient luck
the firm that first gets off to a good start takes the market this is well
before Google by the way our Facebook and then he gives up and then fast
forward to my hero you also have shrimp at our the patron
saint I hope of the complexity hub in Vienna I
don’t know if he was an early member but anyway
Schumpeter who’s writing a history of economic thought doesn’t like multiple
equilibria you know well the outcome could be this however it could be that
or it could be this and he remarks that multiple equilibria or may exist he says
but we can’t really allow them because they would produce in his words a chaos
that is not under analytical control economists didn’t know what to do
if you allow positive feedbacks could go this way small deviations go amplified
this locks and could go that way what are you going to do could so there might
be multiple so in fact it’s very easy here’s a question
if a standard like typewriter keyboard the arrangement of letters on the
keyboard were sufficiently prevalent that’s you’d have to learn that standard
well hang on have you learned that standard the the typewriters of
everybody knows the standards prevalent typewriter is going to be produced in
that’s standard typewriter teachers will know that standard
well how many such keyboards would there be if you count the number of letters on
a keyboard I think I’ve done this it could be anyway I remember counting up
all the roses about fifty four well so in principle we could have fifty four
factorial keyboards just Auraria rearranging that’s a lot of equity Bria
and I have I have a lot of funds well there could be this outcome but you
could have that outcome but luckily there’s never many are you joking you
know think of all the ways that English could have all the different words all
the different borrowings etc so I was getting interested in this but again
like this earlier problem I was telling you in control theory my thesis I
couldn’t solve it I don’t know what to do economists were beginning to
recognize that realistically there wasn’t just one solution or one
equilibrium like economists think of all these forces as being balanced so
nobody’s any incentive to deviate that’s an equilibrium economists were saying
all these forces could balance here or they could equally balance there and
this was becoming embarrassing well I didn’t know what to do about it
and I began to notice as I was reading economic development studying that this
was recurring and I asked my professor Israel what if
they’re not diminishing or turns on on the March all of their increasing well
then yes let’s go on to the next theorem and as will assume diminishing returns
so it’s like as there can only be one outcome for a ball and a cornflakes bowl
if you assume the cornflakes Bell is shaped like a cornflakes poor whatever
it isn’t ball could end up many different places I don’t know what to do
about this and then I started to read a lot of biology in 1979 I was working in
think-tank in Vienna called Yassa the International Institute for applied
systems analysis and I managed to get myself seconded to the east-west Center
in Honolulu I back to Hawaii and I picked up a book in California along the
way by Horace Freeland Judson was called the eighth day of creation big thick
book it’s the history of molecular biology and particularly it talks about
Jacques Monod and mano and his friend Jacob false while Jacob and Jack Ma the
people who figured out gene expression and the whole timing of that more mano
had written a monograph called in English chance and necessity pointing
out that some small things pushed you this way or that way and then they would
log in I’m going Oh God theoretical biologists know all about
this amazing Wow but they weren’t telling me or anyone
else how to solve this problem of multiple equilibria then I started to
read some nonlinear theoretical physics probably nothing with that word
challenge some of you it is by Herman Hakan
when I was at Yassa and he was talking about the physics of lasers you get
something called mode locking and lasers it could lock into this mode or that
mode or another mode I can see you know plenty about this and what difference
does that make well they had the mathematics of dealing
with this and then somebody told me about a professor in Brussels called
Ilya Prigogine and gave me one of his papers to read Prigogine was not held in
high regard these days but provision had the right ideas I read this in June 1969
I read this essay about prigogine where he says Oh systems can lock into this
pattern or that pattern another I began to realize that you could start an
island this is one of my thought experiment examples I’d been in Kauai
and in Hawaii one of the islands when the first traffic light was installed
and so that led to this following thought experiment what if there were no
cars in that island and Milton Friedman is in charge of the island that may not
mean much to you a great libertarian economist Milton Friedman’s the governor
of the island and he says you are free to choose as a driver which side you’ll
drive on and there’s no bias as all the steering wheels are in the middle so you
come back six months later what you see you see tons of wreckage but you also
see in that island all the cars have lined up on the left and reproduce
something in a different Island Milton Friedman’s wife is running she was also
similar and cars have all lined up on the right nobody forced this it’s just
that if you’re if more people are driving on the left by chance
early on then you’re gonna have to go along with that if more people drive on
the right you’re gonna get killed if you keep driving on the left so people are
free to choose but they start to line up so I read prigogine I read Jack mono and
I had this thought experiment and I suddenly realized I had a way to deal
with increasing returns problems or positive feedbacks my idea was to slow
the whole thing down and look at it step by step and rather than think we can get
this equilibrium we can get that equilibrium and it’s indeterminate which
side of the road emerges why not look at the whole thing step by step as a random
process that’s normally near so I did that so I thought well if it’s a random
process a nonlinear I can apply stochastic process theory if you’re not
familiar with that just think of that as the equivalent in dynamics of
probability theory so what I decided to do is I would model all the small events
in one of these problems one person starts a typewriter another one
somewhere else starts a different keyboard typewriter another one starts
something else and then people start event by event to adopt those
typewriters and under small random events if you allow some probability
different types of people showing up somebody talking to someone on an
aircraft someone having different tastes maybe the government putting its hand on
the scales you might get different small events and all I had to do and this was
all I had to do first to learn enough nonlinear probability theory and not in
other words nonlinear stochastic process theory I could nail this whole thing
I couldn’t see how to do that I didn’t know enough probability theory I shared
an office with what we then called a Soviet colleague Yuri or Molly F they’re
a wonderful man and I asked emolia if I said what do you know about nonlinear
stochastic processes I said I’ve got this problem could we put it in that
form Brian not possible so nothing happens I started to study like crazy
non-linear if you want to crack anything first comes the idea
then comes the technique in fact it was there Molly have taught me that Brian
great composer first comes theme of symphony after
orchestration so you have to write for 35 different instruments etc etc but the
theme according to is the error Mallya theory of science first comes the the
theme my theme was Oh small events stochastic process nonlinear but how do
I get that to work ask our Molly of their most not possible I think a year
went by or maybe several months sitting at my desk thermo lives at the other
desk I can still see this in the Schloss at yasser Ryan
you still interested in problem with multiple equilibria I said yes I am Yuri
Ryan maybe we can solve this because board writes down theory he was from the
score a hard school of probability theory I even met the great scorer hard
Monday in Ukraine they knew how to do probability theory so her MOLLE of
pointed out the key mechanisms even the key textbook luckily it was
Russian but it was translated into English I learned the whole thing and we
cracked it before that I had written a paper at Yassa
with the title of competing technologies so these would be several different
technologies like typewriter keyboards competing technologies increasing
returns and lock-in by small historical events and I hadn’t used any fancy
probability theory just random walk theory which is probabilistic
and that was 1983 I brought the paper ID as a Yassa
working paper and I sent it to the American Economic Review I wrote this
paper and I sort of beheld it afterwards I don’t know where the paper had come
from but I wrote the thing it was single authored I thought this is neat it was
showing that under diminishing returns everything would be balanced and there’d
be a single equilibria under increasing returns which I could twiddle with the
parameter you’d bounce around in the middle but if you went off like in the
bowling alley if you went off to the side game over
one technology would lock in but you didn’t know in advance which one it
would be so I’m very proud of the paper I sent to the American Economic Review
and I thought this is a slam dunk I’ve solved a big problem in economics the
increasing returns problem came back terribly sorry we can’t find any
technical faults in your reasoning however this is not economics I
I sent it to the quarterly review court q je quarterly Journal of economics
Harvard terribly sorry it looks to be rigorously done and proper and but we
don’t recognize this as an economic problem or indeed as economic theory I
sent it to the economic journal similar review I sent it back to the aar because
the editor had changed and I thought he didn’t like it I’ll send to the new
editor comes back six years went by still no publication I wrote the paper
in 1983 and finally by that time the paper was getting zombies that
circulation if you know what that expression is it’s it just means the
paper had gone underground all the key theorists knew about it many of them
including Quintero liked it and the paper was sort of getting this damn it
type of reception dammit why isn’t this in a journal dammit why do we have to
keep citing a working paper so I sent it back to the economic journal in England
with a couple of dammit letters no dammit why isn’t this thing published so
by popular acclaim they published it in 1989 only six years after it had been
written well it’s okay I’m they I want to go we did start a bit later so so the
paper finally got published but along the way I give a talk on the paper at
Harvard and somebody there I won’t say who jumped up and said this what you’re
saying is it’s a failed experiment but it can’t really exist I said why not
well with perfect insurance more that’s what perfect this market was
perfect that market you’ve got a single equilibrium that would be perfectly
efficient well I thought well you know if there were perfect ferries around and
if pigs were truly able to fly maybe the world would be wonderful
six months later I gave exactly the same talk in Moscow at semi the central
economic mathematical Institute in Moscow and I said therefore you can lock
in to one of many solutions with small events determining which you log into
and they need the where you wind up may not be the best possible outcome we all
speak English by increasing returns my ancestors didn’t I bet your ancestors
didn’t either but we all speak English of small events in history actually in
Europe had been slightly different if you want to make bets and the 1500s you
might have said we’ll all end up speaking Latin as a common language 1700
spools end up speaking French eighteen hundreds
a lot of the center of Europe spoke German all right we’ll speak German and
then finally this language from some tiny insignificant island which I can
say being Irish tiny insignificant Island takes over all the entire world
and story’s not over yet it might be Mandarin in 104 indeed in 50 years time
but so small events can bring you into something like that but who’s to saying
this is optimal or any other outcomes optimal well anyway the my paper kept
getting turned down not because people were mean although I’m willing to that
could be a hypothesis but I think it kept getting turned down because Western
journals particular American ones I was basically saying if you leave a free
capitalist market to itself you might not end up in the best possible way
took the same story to Russia and hand shoots up in the audience at the end bloody mere me calif who had a order of
lenin pain very very nice one professor this cannot happen in soviet union only
in capitalist countries so I said well right why not because we have superior
socialist planning and we always end up with correct solution I rushed back to
Stanford put a PhD student under the problem right away he wiped a lot it’s
not true even with a lot of look ahead there’s no way you can say with any
technology or anything we’re gonna get it right because you don’t know how that
technology will improve over time or whether it’s got a lot of possibility
anyway so super obstacles old way and then a colleague of mine before I could
get my paper into the economic journal Paul David who was chair of the
Economics Department produced his own version of the same story he’d been
thinking along those lines himself but suddenly that created a sensation he
wrote it in English prose and arguing history was important it was a rather
lovely paper he wrote but that took the wind totally out of my sails and I think
it led to a whole idea that by Tom my paper came out I had taken somebody
else’s paul davids ideas and mathematize them so this was not an easy time I did
have backers though Kenneth arrow was solidly behind me martin shubik these
are superb economists and quite a few others were saying this stuff needs to
be published it should be published and it’s important
I want to just say where the tide turned it’ll take me another five minutes or so
yeah I got the paper published in 1989 in the economics journal if I look at
citations for that paper five in the first year maybe thirteen three years
later so the paper went unnoticed but it’s picking up in citations it’s just
gone over 10,000 citations and it’s getting about two citations per day I
say that with great Glee ha ha ha but in the strength of that paper I was asked
or got to publish a version in Scientific American and on the strength
of that the idea which still wasn’t accepted in 1990 and economics fully Wow God accepted in the Silicon Valley and I
was asked to give a talk at Sun Microsystems on positive feedbacks
increasing returns what we now called network effects and so I went down to
Sun I thought there’d be 30 engineers there with pocket protectors and you
know rather uptight guys a couple of decades earlier they might have its
slide rules maybe 30 people maybe 40 when I could down there there was
something like 800 and they’d read the Scientific American article and among
them was a guy called Eric Schmidt who wound up running Google and several
other luminaries bill joy his role in the audience I charge a very small
amount of money for this and later Eric Schmidt took me out to lunch I said I
said what do you want to talk to me about Eric he was before Google des
oh well I just wanted to let you know we launched Java based on your ideas and
we’re what we want to know what we can do for you in return I should have said
options and stupidly I didn’t I said look Sun Microsystems has SPARC machines
give me the high highest computer high-end so they did
I got a spark too I don’t know if this means anything to didn’t mean anything
to me I couldn’t program the damn fear and
eventually I just gave it to the Santa Fe Institute as tax deduction and and
there were nerds SFI who could program this thing I couldn’t
and similarly I was anyway there were a lot of Andy Grove wanted to meet me
Intel wanted me to give 10 lectures to their top people etc I suddenly became a
hero in Silicon Valley because what I was basically saying under increasing
returns if you can maneuver our engineer an advantage early on you will build up
a user base if you’ve built up a user base you can lock in the whole market
and so the result is Google or Facebook Google’s competing as a search engine
early on with Alta Vista and with Netscape there’s another one as well
there’s yes and Facebook’s competing with MySpace etc you know you can see
how nerdy I am I don’t really even remember these but if you can get enough
advantage early on you got a user base so we’ve got a user base if enough
people are using Facebook I remember asking my son quite a long time ago 15
years ago or something I said myspace is gonna take over and he said no no I said
well what do you think he says no no it’s gonna be Facebook
I said why he said because that’s what all my friends are using so these are
unstable markets the more people use Facebook the more you won’t need to use
Facebook similarly with filler the larger
companies there’s increasing returns I had been told time after time in these
reviewers and by colleagues at Stanford this increasing returns work is
interesting but it’s kind of like black holes in 1930s we don’t think it really
exists well as typewriter keyboards or languages okay but these are all trivial
examples they don’t matter these are curiosa they’re trivialities I
was walking down the street I remember from a car park to teach the very first
summer school at in Santa Fe Institute in 1988 and it occurred to me that as I
walking down oh I thought oh my god all of high tech all we don’t use the word
high these days all of tech our high tech this 1988 runs according to
increasing returns oh my god I wrote that in a Harvard Business Review
article the one that Steve mentioned 1996 the one that was edited by and
rewritten by Cormac McCarthy so I tell I’m sure you know Cormac honor yeah he
was my office mate and we spent four days rewriting that article what Steve
didn’t tell you was I was boasting about this at the Santa fans Joe that this
famous author had rewritten my piece or shown me how to rewrite it and the word
go through to the Harvard Business Review people my editor calls me she
says I I heard your pieces got edited and rewritten
I said yep by Cormac McCarthy I said yeah
what did he do to it I said well now starts with two young guys in Texas on
horses who discover increasing returns so anyway let me finish up here and the
idea not only became popular in Silicon Valley but there was a tech frenzy in
1999 everybody in their uncle was doing startups and they were doing these they
were and they were arguing in front of venture capitalists well we know that if
get enough user base you’re gonna come out on top and dominate the whole market
we don’t look very impressive but we’re gonna go after user base so just give us
50 million dollars and we grab all the user base we can and we will dominate
the market we’re gonna be the next Microsoft or you wouldn’t have sent
Facebook quite then we’ll be the next whatever like Google the problem is they
might have 19 other rivals who’ll be saying the same thing and 19 people or
19 firms are gonna lose one is gonna come out on top how’s MySpace doing have
you been on that recently what about Alta Vista have you looked at
that recently yet cetera but that was all forgotten there was a frenzy and in
the press I started get blamed that I had fanned the flames with all these
increasing returns hopes I wrote a little piece saying if you’re an
increasing returns market the chance is already going to lose but you might not
lose and so I started to talk about what I called the halls of production
diminishing returns economy and the casino of Technology
I imagined an each technology game that was starting up like search engines
would be a table you go up to the croupier at the table and you sit at the
table is that I’m going to play this game of search engines whatever the year
was 1995 or something on the croupier has always frenched the capiases
oh you’re very welcome to join us and you say okay does cost anything to ante
up the croupiers as obvious you can ante up that will be four billion dollars
please so you ante up and say what are the rules of this game we do not know we
do not know what the game will be okay okay well I’m still in I think I’m
pretty smart what is who’s going to show up and play
and the creepiest is we do not know and do you think would be government
regulations the crew be is’s sir do you still want to play this game but that’s
as I call that the casino of technology and my god that was Silicon Valley right
there and so everybody now talks about network effects and lock-in I went to a
McKinsey I don’t know teaching session actually an Amsterdam and the teaching
session guy was teaching Network effects and lock-in I walked into the back of
the audience systems but when I first knew young and I walked in the back of
the audience and this guy was really good he explained it better than I could
and but my name was never mentioned at the very end I thought I’m going to take
advantage of this I walked up and I said thanks very much I said your colleagues
said I could sit at the back he says oh not at all who are you I said I’m Brian
Arthur he turned white so the last thing I want to say is that yeah things have
come out this is now in textbooks is now taken for granted but the moral I want
to draw from this particular session just to finish it is to say that what
sort of economy are we looking at here if you believe there’s lots of
positive feedbacks as well as negative feedbacks it’s an economy that’s not
necessarily optimal we we could have locked into inferior technologies and
indeed we did it’s an economy where small historical events no Queen Anne in
England liked T before that people in England drank coffee this might have
been around 1805 you know and thou great Anna whom 3bet realms were bad off
sometimes Council take and sometimes tey she drank Chinese tea England locks into
tea somebody more fashionable by the 1960’s drinks coffee now we all drink
coffee mr. starbuck or whoever it was that in the vent at that anyway we’re in
an economy that I would say is like anybody’s life you do something you lock
in their historical circumstances you take a job in Singapore you might take
up and marry somebody whatever it is your life is historically dependent if
you’re you might say it’s known our ghatak a few like that expression
there’s no single path non-air god or air goddess and but you’re locked in too
many pathways and they’re always branching and one thing leads to another
so this is already quite far away from the berkeley scenario I was describing
you earlier where there’s one big cornflakes bowl and the economy is
always uniquely in one place and there it doesn’t matter how you got there
history doesn’t count and that’s perfectly efficient we are not in an
efficient system there’s no way that my life is sufficient it’s one circumstance
after another I met somebody a professor of minors
I said this was in 1969 I said I’ve I’ve applied for twelve jobs for the summer
can’t get it can’t get it accepted in any of them was during the Vietnam War I
said I don’t have clearance to work at Rand I’m not an American etc he said
well let me call Dave Hertz in McKinsey and company do you speak any languages I
said I can speak some German and and some French that was good enough
McKenzie interviewed me in New York in German for six hours and my life was
different after that that’s probably a good place to quit thanks for your
patience young and thank you thank you Brian
the we started a little late so we finished a little late and we finished a
little later again but that’s okay we still have about 20 minutes or 25
minutes for questions and discussions with Brian who would raise his fingers
for us just go ahead thank you very much for the for the talk yeah I especially
enjoyed going through a history lesson of all the things that happened before
me and in the context to how the ideas emerge I’m just thinking that
governments around the world trying to create Silicon Valley’s of their own in
their own country given the path dependent yeah argument that you made
are they bound to fail given that the fields already structured
sure yeah if I had one if I relied on answering three letters I would say yes
I’ll give you a lengthier answer and if governments are trying to set up their
own Silicon Valley homegrown is that bound to fail the answer is a
little bit more complicated I was writing you’ll hear in the fourth talk
I’m giving about technology but I made a very detailed study of Silicon Valley in
fact I’ve been in Stanford since 1982 so I’ve seen close to 40 years worth of
Silicon Valley Silicon Valley didn’t start in the 1950s or 40s as I was told
hewlett-packard it started in 1907 with a firm called Pacific Telegraph and that
firm started promising to do wireless Telegraph for u.s. naval ships offshore
and the first venture capitalist in that firm was Stanford University Stanford
gave quite a bit of money Fred Terman was connected with them in due course
Shockley who invented this semi conductor at Bell Labs had been a summer
intern there etc the guy who invented triode vacuum tubes lee deforest came
out to help and eventually a continuous wave amplifier did come out I’m not sure
it was from this particular firm the point is that it took about 50 or 60
years of organic growth to get Silicon Valley
why is Silicon Valley where it is in fact I used to get caught up in this
question emotionally why isn’t it Stanford why isn’t it it Berkeley I mean
Berkeley was a top dog long before Stanford the answer is that
it could have been almost anywhere it could have been possibly in Oregon but
what fed Silicon Valley Stanford University was open to their
guys to their alums setting up firms Fred Terman was the Provost of Stanford
he gave lab space to his pupils who names like Hewlett and Packard grad
students these were the Sergey Brin’s of the day they set up companies
hewlett-packard and other companies and the whole thing and this is what I want
to stress Silicon Valley grew organically so if you think in terms of
you can read journal articles you can read engineering magazines and you can
set up a Silicon Valley I went to Finland to see if that was possible
I’ve sent there by the world bank not easy what is organic about it so think
it’s more like getting a rock garden to work or a little like ecology and IKEA
logical park and part of Africa you can water it if it’s a rock garden or you
can weed out things you can see what’s taking but the the thing about Silicon
Valley’s it’s about people the expertise resides in their head if you could get
all those people and transport them wherever maybe to Singapore in the 1950s
and precisely those people and people similarly trend you might be able to
plant as Silicon Valley here I don’t know but I think it’s all about people
and in the end high tech is not science high tech is what I call deep craft it’s
knowing what to do it’s knowing what not to do for the last 20 years I’ve been
with Xerox PARC that’s where I work in the system sciences lab there if you
don’t know something you could down the corridors I need to do some edging 3
microns because I’m laying down a circle circuit and do you know anyone who does
oh you have Fred down the corridor knows about that you go to Fred and say here’s
what I’m planning to do and Fred says I don’t ever work why we tried
that 10 years ago let me tell you what dolls work so this is the whole craft
thing and that’s very much Silicon Valley very hard to plant that and
Silicon Valley’s full of spin-offs I don’t know if it’s like a mangrove swamp
or something like that but Shockley’s basic transistor team didn’t get along
with Shockley so they walked out very famously and they started Fairchild
Semiconductor and people left that and started Intel and people left that so
these are all little pods growing the whole things organic and the main
message I want to give you in this talk and the third one is that the economy
isn’t a machine you can’t easily start it up it’s organic you can transplant
stuff you et cetera I don’t know the nearest thing I’ve seen to Silicon
Valley’s in China Gendron it’s a fantastic place I don’t
quite know if they’ve got it but there’s the energies there and it
may take off and there are other hopes Singapore is by no means elementary and
the Singapore’s got really to top universities I’ve forgotten the name of
the other one but it’s pretty good and it has top engineering and that’s really
how you grow these things you get people who knows stuff but the stuff may not
even be in textbooks you know ask yourself if you if you want to
understand how you get a Silicon Valley where did rock and roll come from where
where did jobs come from where did certain type of poetry come from it’s
the same process very hard to do top-down in fact I if Steve doesn’t mind
if I tell a story I’m sitting in the president’s office and NTU
about six years ago and someone says in fact it was su Guan Ying the president
of and he only says all right we’re gonna start this complexity Institute
where do the ISIS but where are we going to get the ideas and the projects from
so I sort of hand wavey oh I said all that will come bottom-up you know from
people like you postdocs predock’s smart people even
maybe professors so all of this is going to come bottom-up dead silence in the
room and because bottom-up and I hope you don’t take it as critical but it’s
not exactly a Confucian concept Lee delicately say so there’s silence and
then finally grinding breaks through the silent sisters yes he said bottom-up we
can arrange that another okay next question or comment thank you I actually wanted to come back
to a question you actually asked herself at the very beginning of the talk I
think you said and I paraphrase what I wish I knew when I was 24 thirty years
old I don’t think if I’m not if I’m right I don’t think you addressed that
touch back on that okay hmm I planted this guy in the audience I was
watching last night Lord of the Rings where Frodo starts off and says I’m
going to take the ring into Mordor that’s what I learned I learned that you
can you think that all of this is gonna be easy it’s gonna be simple but you
know there’s big spiders and stuff along the way I I thought that if you one key
thing I’ve learned and I’m sure if you’re a bit older you know this
automatically in science I didn’t I thought if you had a very good idea
bells would ring and you’d be carried in in people’s shoulders and so on no the
main thing I learned is that there’s pacing to science and at first if an
idea is sufficiently different it’s regarded as you know it’s okay but you
know this isn’t fill in the blank this isn’t obstetrics this is not what
linguistics and I’ve seen so many examples of that I remember Greenstein
at Stanford and formulated a very different way of doing linguistics the
ancestry of languages and I hope I got his name right and for decades he was an
outcast it turns out that he was very much on target so unfamiliarity and lack
of say you know the question I thought I’d solved and I think I did solve was
how do you deal with increasing returns problems well you may you make them into
stochastic processes you said small events are gonna sway the outcome then
you can well define them but when I said that journals nobody heard of the
increasing returns problem so I’d solved a problem that nobody had written down
and I’m not saying and I’ll give you other examples as we go along I wish I
had known that now I do and if you’re wondering where you can really learn
about doing science and about careers I would say read and read and read
biographies it was after all of this I had hair that was color of yours for
this story I went gray trying to get this thing published I think that I just
wish I had studied more biographies early on and the enemy is not people’s
stupidity it’s not at all it’s it’s a it’s unfamiliarity saying well this
isn’t really a problem I mean what you’re trying to do here and
occasionally people would say I get what you’re trying to do but I don’t like
that direction you’re trying to take economics in Stefan Stefan and you I am
NOT a scientist or an economist so this might be a naive question but I was
wondering if things are not or systems are not as predictable as we think they
are and I can see like this I think it’s very beautiful to see the world in an
organic way and thing small things affecting each other contingently but predicting or being able to predict
things in a deterministic Newtonian world
have been able to allow us to base or to create models through which we can
understand the world so if we look at the world in a in this kind of organic
way then how do we then create models off of that so how how is that possible
yeah these are all excellent questions my answer to that would be the world
isn’t organic doesn’t mean it’s chaotic it doesn’t mean you can’t predict
anything the world is and I find this comforting world is partially
predictable we can predict the orbits of planets very well if you’re very good at
psychology you may be able to predict other people how they’re going to react
we can predict an awful lot but the predictions are not perfect there was
definitely view in the 1700s like around 1750 that the world was ordered and if
we knew the mechanisms well enough we could predict everything now we’re
finding out that if there are nonlinear systems earthquake systems for example
this triggers is triggers that might trigger this or might not then the world
isn’t totally predictable so I’m more comfortable saying that things are
reasonably well ordered things are some some aspects of things are predictable
and some aren’t and I think that for me that gives more of a world that I like
history matters you know Steve Jobs hadn’t talked to somebody called Wozniak
and they hadn’t tried to do something I’ll give you this example Jobs and
Wozniak visited where I work Xerox PARC one day
there were two managers decided whether they’d allow the visit one of them was a
male and said yeah they can come and visit there no they they’re just a
couple of kids who were hobbyists the other one was smarter the woman and she
said they’re gonna come in and we let them romp around Park they’re gonna
steal our technology but they let them in and they saw the I think was the Alta
parks computer and Jobs and Wozniak went back and we reverse engineered that and
produced something they called the Macintosh and that was a knockoff of
park anyway the point is that these are tiny events and Apple is now I haven’t
checked recently but I think it’s the largest company in the world and if they
hadn’t had that knockoff it would be different so I like I like that it’s not
totally predictable it’s biology working differently in some or is it is it also
working along the same lines of increasing returns I think that there
let me put it this way I think I bet you you could answer that better than I can
because there are people in this audience know a lot more biology but there I probably think of a good answer
for you in two days time what I would say at the moment is that any field you
take it could be some field in physics or in in epidemiology or something
generally those systems have a mixture of increasing returns or positive
feedbacks and mixture of negative feedbacks in the case of biology just
tech species there are many cases in which species
once they get established they’re very hard to invade vaguely species of carp
that were in I don’t know if it’s car but introduced into what is it Lake
Tanganyika it’s my own ignorance here but you introduce one species of fish
it’s wiped out all the others so you see a lot of this in biology and there’s
something biology called competitive exclusion if you have enough of one
species it’ll wipe out anything that’s similar to it I don’t know to what
degree that holds up empirically but I would say this that’s the mixture of
positive and negative feedbacks is all through biology genetic regulatory
networks etc but particularly in competitions among species in fact I got
these ideas from biology it was Jack Mono saying if you have positive
feedbacks hole patterns could get locked in that the molecular level once you
have a replication mechanism which goes way way back in time in biology maybe
four billion years or more there’s no other mechanisms gonna find it easy to
get a a toe in so we’re full of these historical accidents that lock in that
creates the structure for them the next set of things to happen I thank so much
for this my question actually builds on the lady in France writes about decision
making so if there are 54 factorial possible you know permutations of a
given outcome how do you make decisions let’s say if you’re in government and
you know you need to make some kind of policy that will affect a lot of people
how do you go about doing that in a way that’s I don’t know a better I don’t
know a better word than practical no that’s a terrible word that’s consonant
with what you just heard I do how do you make decisions at the
governmental level of two or three observations one is that if you know
that these systems can lock in that some standard may lock in
you can gently nudge the system you know if it’s teetering on the boundary and
you say oh it could go this way for example there’s you can get into a
cycle of positive feedback with name-calling between America and Russia
at the moment well possibly you can nudge the system in in a ways into a
better cycle of solutions also I I’ve come away I was sort of taught that
economies and these decisions economies could be looked on as gigantic like
electrical power systems or something you know these big dials and levers and
behind some glass wall all of this stuff is happening but in the Singapore
government you can pull this lever that after and you’ll hear more in the third
talk but after thinking of it that way I yeah to some degree you can maybe
control the roads shift system and traffic maybe that way Singapore does a
very good job of that it seems or a reasonably good one but I tend to think
that economies are more like I’m trying to think of the word there they’re more
like nature reserves you can put fences here you can bring in
new species you can replant here you can wall off things you can encourage this
or discourage that bit like rock gardens and you can sort of organically replant
and fence off but it’s very imperfect and I think the and also if you were to
say to me well this is all you know I just heard from this talk that the
capitalist system doesn’t lead to the best of all possible worlds review leave
it to itself all kinds of things happen and we lock into large you know large
companies Facebook’s one where the last couple of weeks we haven’t been that
pleased with the outcome I would say the kind of to first order capitalist
systems do deliver but you can’t say they’re perfect and America doesn’t at
least twenty years ago during the Cold War 20-30 years ago it did not like that
thought we are superior we are the best and the amusing thing was that the
Soviet system thought the same I think planned planned capitalism a la China is
doing very well hands-off capitalism which is more in
the West not perfectly hands-off is doing pretty well too so the economy
does tend to reasonably good solutions but get away from thinking this is
perfect and governments can intervene here and intervene there but if you
follow this sort of reason you need good timing where is something just teetering
on the edge because it might be too late it’s too late now for Singapore to say
we don’t want to use English as a common language but it could have said that
maybe fifty years ago I don’t know could have said well we’re going to use
heavily go Chinese and let English gradually die I don’t know the timing
counts for a lot any other questions if not then I think
we should thank Brian once again thank you thank you very much Brian once again
no thank you and join us for coffee

Leave a Reply

Your email address will not be published. Required fields are marked *