Alan Kay talk at UCLA CS Connection Lab

From Viewpoints Intelligent Archive
Jump to: navigation, search
okay everyone um
I'm here today with isan K we're at the UCLA
connection lab uh with professor kleinrock and the rest
of our students here uh we are very lucky to have Alan
who is presenting on a broad range of
topics ranging from technology society and future research
and I'll keep it short so
uh Isen needs no further introduction I will let him begin right
away thank you so
much okay onward we
go so the way this started off is my
long long friend uh
Len kleinrock sent me an email
saying would I do a talk
and I wasn't
says about the view of the current you know basically
about everything most rewarding
work in my life I'm 84
years old guys
and can I do it in just a few
minutes and of course I said okay
and uh the wonderful
Jason gathered
up a bunch of questions and
I endeavored to look at all all of them
and combine them in one way or another
but uh
the short an answers here
are uh is it
still feasible today to do something like zorox Park and
the answer is yes I don't know of
anything like zorox Park in the
computer realm of things
uh they're all so goal
oriented uh but I've there's
a great Biology Institute funded by the
foundation called janelia labs
in near Washington DC and there's the famous
Max plank Institute and
see a little bit why I don't
consider there's a lot with
the same kind of spirit as ZX
as I'll say in the moment it wasn't zerox
Park zerox Park was
part of the much larger
arpa uh research
committee and considered it it was just funded by
Xerox because the department def had quit
funding arpa the right way and put the D on and changed
everything so we'll see about that
in a minute and
a short answer do I believe PE young people today
have to solve a harder problem in requiring
resources and my answer is yes
absolutely it is much much harder
as the businesses
have moved in
which started really strongly in the
80s after the inventions of personal Computing
and uh P pervasive and local
networking there was something to
that could be and
the and the gooey which
open up the computer to the
sorry go ahead I I could hear people talking
yeah um yes so
it's much harder much much harder
there wasn't time to really answer
these uh completely separately
so I sort of wo them together
and I
like this one this is a recent one what are some
of the most tragic illusions that adults
with that is the most flattering thing I've
heard in years because it assumes I'm not an adult
which I'm not
uh number two was what did you guys being there at Xerox
Park do for them
and then there's this dot dot dot one
which I did do a little prep on it but I
left it to the end in case we run out of
time because in 84 years
uh I've had a lot of aha moments but
um let's start off with number
one this is part
of the tragic Illusions
what we're doing right now and what we do in most
classrooms is to recapitulate
something that happened 150,000 years ago
which is a bunch of people gathering around a
campfire listening to somebody
talk and just
think about it for a second how crazy
is it for me to be doing what I'm doing now
this is why I sent a couple of papers along to read
it's crazy to orally discuss
ideas inventions that required reading and writing
and explain the
oral medium is not good for what it
the best of the stuff that we
do but
we all have the same pink brains
they love stories and so
what this guy wants is a story
and what I'm trying to do is
to comply to tell stories about
things that really aren't very good as
stories and the problem was we're still basically
the same uh species
with the same pink brain that wants
tories and systems
aren't stories I'm basically a systems
and if we take a look at what happened going back a few
hundred thousand years human genetics
yeah it's changed a little bit we
some of us got to be able to digest milk
and so forth
and human
ventions gradually took
off and with modern
theories about origins of languages
a lot more of language seems to be an invention
via culture than
uh we thought in the past and then we
have this problem which I hope you can see
there because my little thing is
blocking um
our Collective wisdom
started being informed by our inventions
but it grew at a much slower rate because the
inventions are uh various forms of
technology including symbolic Technologies
they turned out to be really easy once we got the hang of
it and so we've got this
huge gap
so we can turn the guy's spear into a Tomahawk
missile we can increase his power of
Destruction by a factor of a million
and we could do something much much worse which is to
turn him into a bureaucrat
and create bureaucracy
and bureaucracy has killed millions
and millions more people than the worst Atomic
weapons so
this is a huge problem
and education
something about it we'll talk about that a little bit
later but think about it we
just went through the inability of most
of the population of the planet to not
understand how viruses and diseases work
that was botched almost
everywhere and for
the last 60 plus years we've been
ignoring something smaller than a virus the CO2
which is gradually
changing the entire planet that we live
on so if you're looking for
human wisdom it's not up to
the scaling of what we've been able to
okay so cliff notes for human beings
I'll do this quickly
but so we're basically
our minds are like theaters
we're more theatrical than we think we treate
create our treat our beliefs as reality
that is the worst thing about human
beings and these theaters are tiny we can
only think of a few things that once it's hard for us
to take in larger
things and if you look into biology
we find out that most of our mind is not
human and we're more non-human
than we assume and we tend to react
much more than actual thinking thinking actually had
to be invented we grow up in
we were born into
and that shapes so much of who we are and also
how we look at the world our conclusions
tend to be societal that's a
disaster and we have
not just one mind and
really not just one brain we maybe have
about 80
different conglomerations of things that could be
called brainlets
uh we have a special set of ones
shared with most animals which is learning
by touching things we learn
seeing things we have a completely different set of
knowledges from how things look
and uh we have this
symbolic mechanism
which is both our
uh down downfall and our
Salvation so we're much
more fragmented we only have one body
but uh what we want
actually conflicts
internally so this is
a a whole area that's really
worthwhile studying especially in
this age of computation and especially
this area I'll leave it at
that now let's go to the the
second question is what did we ever do
Xerox anyway
well we'll talk a little bit later about where zor
Park came from but here are some of the things that
did so the uh first
thing like a modern personal
computer uh the the
goey with icons and
basically a way of
uh dealing with symmetrical
reading and writing that was also what you see
get what we have to call real oop today
that we only called oop back then but it was
a different than what is called to
today and we did the laser
printer which just for the engineers in the
crowd the very first laser printer was a page
a second which most people have never
experien today at 500 pixels
to the inch so that is a great
story all by itself How Gary Starkweather
did that uh
the outline fonts and postcript the ethernet
peer peer and client server and
about half the internet
actually had its own internet which is called the park
Universal packets uh
it was spread all over the country and
uh about I don't know
half the general meetings
that were held at UCLA especially
were done by Park people some of whom
had already done some of the things that working at
yeah the thing is is that
all of these things came out of a particular
community so we locked the claim we
had eight and a half visible inventions and
of course what's interesting about all of this
is all the
technology that had to be invented that was uh
invisible and according to Butler
Lamson this was done
uh with about 25 researchers over about five
years these particular ones I'm showing
here and the park
park itself existed in the form it was for
about 10 years so in
total uh today's dollars
here is like I don't know 15
years ago or so when I made the slot
so it was uh in
dollars of 15 years ago it was about 10
million dollars a year or about 100
million to 120 million for the
for the 10 years and park had four Labs
so there was system science computer science
Optical science and physical science
so this is the total budget of
Park and when I
first did this slide of 15
years years ago or so the return was about $35
trillion to the world now
I think it's up past 50 trillion
or so and to get back to the question
is uh
oh and this is an important idea that the thing that confused
people and Confused Xerox was if you can
these eight and a half inventions you see in
in Industry it is the
basis of the one of the main
large Industries uh that we have
now what about return to park well Park
missed I mean zerks missed most of these
but it doesn't matter because the return
on the laser printer all by itself
was about a factor of 200
on the entire investment for park over the
ntire 10 years so
Zer made billions and billions just
from this one thing they missed the
trillions but
uh if you look at it from a positive direction
thank goodness they're willing to fund
people that no business person in their right mind would
fund so we got what we
got now
the real story here is that back in the 40s
the Cold War started with the Russian AB
bomb after World War II and that
gave rise to a bunch
of DOD funding which included Whirlwind
which we'll talk about in a minute the
sage air defense system
we'll talk about in a minute and then in 57 Sputnik
came along and that created arpa
we'll talk about that in a minute and here are some
the things that arpa did
up to the time when the Vietnam War and
Congress decided to kill arpa funding
for public domain
Computing so you can look at these yourself
so things like uh personal Computing
uh computer Graphics
packet switching that handsome guy in the patch packing
switch switching thing on the the right is
none other than Len
kleinrock looking very confident as we all were
back then that led to the arpanet
and that led to the internet
and the first tablet was really
good one that was done back in 1964 and
then there was the engelbart system and so
forth so park was
actually set up
because uh one of the
uh Deputy directors of
ipto feared that this work was not
going to be continued to some kind of fruition
so he started looking around and it happened that Xerox
was looking around for a long range research
lab and they got together and the
the lab was put as far away from corporate
headquarters as possible uh
next to a university so it's either going to be Berkeley
or Stanford they put it next to
Stanford and um
Park completely thought of itself as
being part of the Opera Community almost every computer
researcher there had been a graduate
student at and had their degrees
paid for by arpa and so forth so I never
talk about Park without bringing in
larger context and in fact
it the larger context was the reason that Park
was able to do so many things
with so few researchers there was
all of this prior art available
in the research community and the resch search
Community also taught the grad students how to work
together I'll talk about that in a
okay here's uh feedback from
world push back look at some
disly low
like here are the
Heartbreakers on the other hand 100%
of the internet 100% of the ethernet
100% of the laser printer outline
font pierer and all that stuff were accepted
and this is in large part because there weren't
good Solutions already out
here the
50% on the PO modern personal computer
is because the people who adapted
this to the commercial World
basically adapted it
uh in its appearances they didn't
look any deeper uh
most of the hackers in the 80s thought they knew
how to program so their interest in doing
really new ways of programming was very low so the least accepted
thing was the way we did successful
software at Park
and some of the deepest
ideas uh that our field has ever come up with
were done by uh engelbart and his group
and almost none of those have even
be understood and you know if I were there and person
what I always do is ask students
what angelart did and they usually can't tell me
and it's hard to find a student
in fact let's do it right now just for the hell
of it okay students
can you hear me raise your
okay let's see
what do I want here
yes so students
please tell me anybody who's T
typed e n g l B- a
r t into Google
raise your
I think one hand is Professor Clan rock well
of course okay so
this is basically being in
being in physics being
a physics major and uh stop being
Newton so this is not an
exaggeration think about
it and you know if you want some bitching from an old
fart a lot of us back in the
old days put out a lot of
uh effort including Len in the
room to get you guys personal
computers which you have with you
and to get you guys networks that can connect
to things and to get you so you can be
just a few characters and a mouse push
away from satisfying your curiosity
about the heroes of your own profession
think about that and none of you have done
that what do you think about
hat okay
so let's now go quickly back again
Russian a bomb
Whirlwind so
this is just a few racks of
Whirlwind at MIT was
actually just a 16bit
bit parallel real-time mini computer
but it took this amount of space had about 5,000 vacuum
tubes and the memory
devices back around 1950
were so bad for storage
for Ram that
uh it forced uh one of
uh the guy in the middle there
Jay Forester to invent core memory
which was the technology that let computers
uh go commercial for the first
time and the guy to his
uh let's see to his left the
guy who's uh
standing behind them in the middle of the picture
there Jay
whoops oh now it's not letting me
I hate Zoom
okay also with this
was uh one of the earliest displays
and uh Everett had
had uh
sorry ever had
actually invented a way to point at the
screen and this is the as far as
the first time there was computer
graphics with a pointing
device and they did this because
the Cold War changed their original
theory about what they were doing which was
uh to make flight simulators
uh and they were starting to lose their funding for
that but uh the atomic bomb
of the Russians got the military
interested in whether you could hook a computer
up to radar and track it and they were
the first to do this and that
hey built back then morphed its way into the
worldwide uh airplane
tracking system that we used to
day then the second phase of Whirlwind
was to go
all out and see if we could
track Russian bombers
before they could uh get to us
and so they built these monstrous
block houses out of concrete
this is about the size they
were bottom floor there is just the
power supply the second second floor here
are two really big Q7
computers with maybe 55,000
vacuum tubes gang
together uh they're really something
walk around in
here's a block house so
24 of these were built in the
US then up on the top
floor were 100 to about
150 graphic terminals
with pointing devices for dealing with various
of information and
here's what one row of them look
like so there are zillions
of them and this is about maybe 1956
and they had a network wasn't
a great Network it wasn't a packet switching
network but they' gotten used to the idea
that if you're going to build computers and have people
work together you better have a network and some
way of doing that so this
crowning Glory was just coming to fruition
when the Russians decided oh
well let's just shoot some
Rockets to show The Americans that we don't need to
deliver bombs by a bomber we can
you make icbms
so this shocked the and
Congress and they set up arpa
Advanced research projects agency
the director uh
in the early 60s was Jack
ruen and
he decided to
uh give this guy uh
Jacob licklider known as lick
some leftover money they had which
by DOD standards wasn't a lot but by computer computer
standards it was millions
and they they gave it to him because they liked him and you'll
you'll get an idea about this
soon and so here are the
1960s directors of
arpa not a military person among them these were all
scientists and the four arpa
computer research directors were all
scientists and this is because
the powers that set this up
realized that having something like the military
and something like congress
with a dire threat
were trying to deal with would just botch things up
still true today
so but
of course Congress ultimately control
the budget but they didn't have
much power because the Americans were were
frightened of the Russians
and so they weren't
going to curtail military spending
but people in Congress would say
these words that everybody has heard Still today
how is this relevant tell me why this work that you're doing is
relevant to the Department of Defense and
they would get push back from these
scientist Jack would say that's not the right
question to ask Bob spra Senior would say the
right question to ask is how is this going to help the United
States or this technology or our society or our
culture generally holy
Bob Taylor's remark
was they would stand up to these guys
and in a polite civilized way attack their
myopia because these arpa directors were scientific
and Bob reflects we've had too few of those people
in the job since then
and around 1966
Charlie hersfield asked bobw what
what do you want and Bob said uh this network
is what I want to do and here's why
and Charlie said
okay and Bob's recollection that was a
15minute conversation
Charlie asked how much money do you need and Bob said well a million
dollars is so it's like seven or eight billion seven or eight
million dollars today just to get it organized
and uh herzfeld said
okay and Bob's recollection was that there was
no actual arpa paperwork order written or anything
for months maybe even a year
these people all trusted each other and
did a lot of things very quickly this
way now a key idea here if you can see
him is that Bob
Taylor uh after
arpa went on to be the
founder uh of the computer science
research at Xerox Park so he was the
connection as of course the computer science
research park at Park was set up like
the computer science research of
arpa and so when we look at
this we want to see
what what did these people weigh back then
think about this we have all
these questions that we can't answer all of them today
we can look at the first couple there and we can get a
couple of samples of what they were saying
so veva
Bush uh pronounced the way
Beaver is pronounced in New England beava
beniva Bush who is president
Rosevelt science advisor said
in the 40s holy new forms of encyclopedias
will appear readymade with a mesh
of associative Trails running through them ready to be
dropped into the mimics and they're Amplified so
that got a lot of people
thinking Li glider said in a few years
the combination of humans and computers will think as no
humans have ever thought
before John McCarthy looked at
the sage system and said every home will have one
and with an intelligent interface because
he thought way McCarthy thought about it is well this
is just like uh getting gas and Water
and Electric power and so forth into a house
it's a utility so every house
will have a uh a terminal connected to a
worldwide utility that was
1958 that affected the work at MIT
by uh Corby Fernando
corbo Ivan suin
the inventor of computer Graphics said a graphic input
coupled to some kind of computation which
is in turn coupled to a graphical output is a
truly powerful tool for Education design
and angelart here said we can boost Mankind's capability
for coping with complex urgent Problems
by augmenting the collective IQ
of groups not just augmenting people
but augmenting
and I'm going to more concentrate more
on these two but we'll get a look at a bit
okay so what what with these guys
looking for as far as people look was just
looking for the best people Ivan was
looking for people uh who could choose their own problems and
methods Taylor's
thought his job was to
fix up the social environment so that when the
no Lone Wolf scientists needed to cooperate
they will and
Larry uh said
everybody understood the game
here which was arpa
ipto was allowed to help the country as long as it was helping
the military and so what
kind of people were they looking for this man right here is
my great-grandfather he's the first cat herder in our
family don't let anybody tell you it's
easy anybody can hurt cattle
holding together 10, half wild SS
that's another thing all together
well you get you get the idea
and so the real question here is what if only cats can do this
stuff and adults don't
get it adults are into command and control
and command and control doesn't work with cats
have a jar fill all the names of the toys we have
sitting right next to us and we're going to take the jar one by
one open up these toys all right would you like to go first
piing something from the jar yes all right
now it's on a spring I like
it's inside this box right here
there's this little mouse on a spring is go up and
down and cats are going to try to catch it all righty our cats
like having mouses but they keep on tearing them apart CU
they're kind of vicious so got another one
okay and this is the toy
mouse what is this
what is this mysterious
George I think you get
it gotta have some feel for what cats
want so
lick pointed out right at
the beginning that great Visions are not the same as
goals but they are the ultimate cat
oy so
he stuck to this Vision which is not
phrased as a goal he said computers are destined
to become interactive intellectual
amplifiers for everyone universally networked
worldwide he did not know
what this meant and in fact he was a psychologist
not a what we call a card carrying
engineer but he
understood a lot
that if you hide if you take this
vision and stick it behind the mountain
range it sends out magnet field which is
Marvin Minsky like to point out if you don't
know what to do get lots of people to do lots of things
that's what lick was doing
and so the number one principle here of
doing anything like this where you are going to
get have the possibility of
really large new things is the goodness
of the results correlates most strongly with the goodness of
the funders that is my
reflection uh for from the past
because every era has really
smart people I've never
been in an era that didn't have people as smart as the people I
worked with back then but the funders
have were wonderful in the 60s
and uh 70s and they've been
uh ever more terrible ever
and so two of the main things that came
out of this was human humans and
computers working together as an idea and
then this larger idea of augmenting human intellect
in general so I'll just go through these quickly
million of them they're in one of the papers I handed
out but basically go against
conventional wisdom do Visions
not goals let
somebody else do the goals fun problem finding
not just problem solving this is a key
one because a problem
is often too rooted in the uh context
he present
and so if you're just doing that which is kind of what's
going on this is what NSF likes to fund
problems that peer reviewers can understand in the present
and most most of them uh don't make
big contributions
problem finding on the other hand is difficult for
everybody but it gives you a
chance fun people not
projects so in other words don't
worry you know if you've got some crazy guy who's really
good or girl who's really good
don't worry about what they want to do
give them some money and see what happens
uh corollary fund of very best people
I don't think that's being done
today and an important part of the research
results are new and better researchers
and in in a sense
NSF is paying for some of this
but I don't think a lot of research is being done I think
it's mostly engineering so what I think is what
NSF is paying for are new and better
uh uh practical
engineers and what we need is new and better
researchers really Advance something very important
well that's controversial because it probably
uh is something
new and here's a key one design
build in the future and bring back to the
present and one that costs a lot of
money was make enough of the inventions
and in a practical enough manner so that
many people can use them in other words live
in your own uh your own mess that
you've made then Milestones
not deadlines
okay so
here are five things that of
many things that people
are in the indidual individual stars there are
people get interested in Aesthetics
or they prefer to Tinker they prefer to do
engineering but the problem is The Sweet
Spot in the center
is something we have to get to because as Marvin says
you don't understand
something if you only understand it one way
this is a key phrase for the 20
20th and 21st centuries
and so no matter what
your favorite one of these is you have to get fluent
in all of them and this is a not
he way universities are set up not the way companies
are set up but you got to do it
the best people I know all are in the center
here and then there's a question about
regarding Computing is is could there actually
be a science of computing they
are artificial objects after all
we're not looking at molecules so
forth we're looking at constructed
things and what does that mean as a
so go back into the 50s
again we've got
phenomena we can see these artifacts they've got
wires and vacuum tubes and we can program them
and we can make a high LEL language for them like Fortran
and so forth and that isn't really
science what is
cience well I've
got this cryptic phrase here scientists try to find
fireflies and make
t-shirts what does that mean
it means that behind the stuff
that distracts us
there are hidden
deeper properties the properties
might not even be in there explicitly but they're
there what scientists do
is good what I call good
woooo which is
they let their brain relax to
point where they can start thinking beneath the surface of
things and here's one of the first computer
scientists John McCarthy and he
looked at this stuff in the 50s and decided to
t-shirt of it and the
t-shirt looks like
this so it's kind of backwell
equations of computing and what's interesting about
it is that it doesn't look like
uh the hardware and it doesn't look like the software it's
actually a mathematical
formulation of what John thought the most important
relationships were so I would call this
maxwellian then
stronger stronger
along the lines of Newton which is to
explain important ideas
here about what was lying beneath
the surface and why it was really useful
to think of doing things in a particular
way and the thing
about math is it has to be consistent meaning
neat but it doesn't have to be
real and so this thing that John did was
not real but another hero
of mine Steve Russell the guy who invented space war
was at MIT and he looked at
John's math and he said that's
a program if I implement it we'll have a running system
and John McCarthy said well he did and we did
and that was the birth of
lisp and here's the cool thing about Computing
the theory can also be the model and the artifact
and you can make it
real but there are a lot of complaints about
dealing with pure semantics is
too slow even after
factor of a million or a billion
of or more of more moris
law still seems to be too slow
on the other hand
you can build a special machine for
these semantics that don't fit on regular
machines because remember
our computer's only there to run programs so why not make
a computer that will run the programs you want to
run and if you try to optimize too early
like buying a computer before
that isn't a supercomputer before you've done your
research you're in trouble because now you're
forced to deal with its problems rather than your
own so this is the kind of process
that the arpa community went
and this is a fun story but
I want to bring up the fireflies again
you can't replicate what John
McCarthy did because the
times are different now but the fireflies
in what he did are still there and they can still
what to do
okay here's uh Story
number two 62
years ago in 1962
this is in the
uh Sage experimental
uh test computer called the
tx2 it was one of the LA one one of these
computers that was the building
and right down at the end of one corner was this guy sitting
of a oscilloscope
screen and what was he doing
here's the light pen he was
and now this display
doesn't isn't a line drawing display it just makes a
DOT so the there's a program
making these lines now he tells sketchpad
to make them all mutually perpendicular and you
just saw it automatically
uh clean up the drawing this is why it's called sketch pad because
you can just draw roughly he may wants
to make a hole in this
flange so he says here make
these both uh uh
parallel and perpendicular and
now the constraint is
make these lines I'm drawing to be collinear with the
guidelines below so this allows
him to quickly draw a dotted
now he's going going to make the guidelines
transparent we get a hole in the flange
and notice this is the first system to have a clipping
window is actually drawing on a canvas
about a third of a mile on a side being windowed
here he needs a
so this is the rough shape of a rivet
draws this because he wants this as
a center for an AR AR for the top of the
rivet and now he says okay make
all these lines I'm pointing to here
make these mutually perpendicular
and that will force the drawing into
something like a rivet maybe he doesn't
because he didn't constrain the size of the thing so here's another
solution with that those starting
conditions but he said decides he likes the one
first now here's another cool idea Ivan
suland did
which was to have Masters and
instances so
this is not the rivet he drew
this is an instance of the rivet he drew
and the rotation and the scaling
are done by knobs that he has on
console okay
and wants to show you
he can make more instances and that
these instances can be different sizes and
says whoops I forgot
I left the crossbow on there I should make these guys transparent
so he goes to the master drawing and makes them
transparent and lo and behold
all of the instances uh
obey so he gets rid of
the thing and he says okay
make that construction I just did into a
master now I'm going to make instances
ight here's
instances of the master
that he made of the
yeah okay
so here's a high resolution picture of Ivan
at the console and he's doing a
something much more complicated which is to
draw in a bridge
and uh unfortunately the movie
for this has been lost it's probably somewhere in Ivan's
garage but uh for fun a couple of years ago
we we uh did a an
emulation of
it so here's a simple trust
bridge and
we can connect things up and if we turn gravity
on which is a constraint
you see it's Computing all the stresses and the
strains on the bridge
and it's continuously Computing so if we add in
another strut
here guess what it ripples over the
entire bridge if we H hang a weight on
it we can stress
bridge and so the big idea here
is it's not just about imitating
drafting it's that we can
simulate uh any kind
of computer AED design drawing that we do
or model and uh
we get not just CAD but CAD
Sim so this
is a huge huge idea
that's still done extremely poorly today
for most
people so
interact this is Ivan's PhD
thesis in 1962
inventing interactive computer Graphics
clipping window pointing iconic IQ objects
Masters and instances we'd call that object
first objectoriented
programming system that I know of
lots of nice technical things in it automatic
Dynamic simulations from
designs and you programmed it not in
the standard uh programming
languages but you programmed it in terms of
goals in terms of the constraints that you want
and sketchpad had three problem solvers that would solve
problems say you could see so I once
asked Ian how could you do this in one year all by yourself
and he said well I didn't know it was
hard check out the fireflies from this
today think about what the fireflies
from this one are okay and also in 1962 we
had this guy who was kind of like Moses opening
he's really interested
in saving the world and trying to make people
better so they can save the world so
his idea in 62 was to
have a screen much he knew about sketch pad he wanted a screen
larger than that he wanted to hang the light pen
from a a gimbal so you're
you could just pull it down to the table and use it like a pencil
and not have the blood run out of your hand
when you're using it and there's this
problem of that he thought about very
deeply which was humans
good Learners
from using tools we
have this pink brain we use tools
brain kind of way
so like a hammer we can use to hammer
a nail into something but we can break something with a
hammer we can do a lot of things with a hammer
and we kind of get a hammerness into
our brain but we get very little
wisdom about tools and
hammers in our brain and
we become a hammerer
100 thousand years later we might decide to
use nuclear weapons on a city that's a way
of hammering them into
submission so part of engelbert's
idea was if you're going to augment human
intellect for crying out loud
you must insert Education and
Training because what's happened is that the
powerful agencies have gone up in power by factors
of millions and our brains
haven't so you want the education
to turn parts of our brain this nice
bluish cast and perhaps give us
a little more wisdom on using our tools and then
a little safer using powerful methods and Powerful languages
and so forth to do things
so this is this five unit thing
engelbart's idea of what you really had to do if you're
going to do this thing and you had to do it not just
for a person but people in groups so the
groups have to be
organized pretty much the same way
this allows to get enormous amounts of synergy
between the individuals they can be
the different stars on the diagram
I just showed you cooperating together
and four years later they
did a first pass at this which was incredibly
impressive it's called the mother of all demos
and uh if you type e n g l
b a r t into
uh Google uh in the first page of
hits you'll find a reference to the movie of that demo
you should see what they were able to do back
little machine
and you know when engelbart's name come up
comes up and even in his obituaries
the the only thing I could really talk about was the mouse
and this really bug
Doug said the mouse the mouse is just
a dial on a car radio we invented a whole
car they won't look at it
you can get an idea of this there are about 70
papers on the website devoted to him
talking about these deep ideas as I
me mentioned previously only about 30%
of what engelbart was talking about
got taken up uh in
Computing today so this is a big huge area
that's got fir Plies everywhere and these are just
three of the 17 or
18 uh principal invest
investigators that arpa put together
so these are the kind of people and
if we talk about cooperation and methods we've
got to talk about the blind
philosophers trying to figure out an elephant
this is a classic thing and
they all have this flaw that I
in the beginning that they think what's between their
ears is uh not belief
but reality so of course this leads to
and after thousands of years the idea of compromise
comes about but this does not
need to a pretty picture of an elephant or a
way to make an elephant and in
fact this is my picture of the way Computing is
today practical Computing is just a
horrendous compromise
almost none of it was
now if you get scientific
philosophers they've learned how to
cooperate in science works by looking at
parts of things and looking
at what other people have looked at other
people looked at other parts of things
and eventually coming to conclusions about the
whole so that's a really good
idea but in fact science
because our brains want to believe rather than think
these can turn into something like a
religion and the reason uh we don't
want to come up with blind belief
there's always
more there's always more and what
it is is something that we can't
imagine so when we give a a name
to something we're hurting our
belief in because we
defined it it's a category now
we've hurt our ability
to think more about it because it already has a
set of properties it already is the thing that the
denotes and here's
a an ad I suggested to Apple
many years ago uh for the
kind of people they wanted to have at Apple what kind should they be and
I suggest well do make a picture of something
like this uh the right
do with
good P implies q and
the left hand the right brain has to be able to do good woow
and what is good wool because
uh the thing the thing to understand here let me just
second here I think to
understand here is it was really hard to
invent good Plies Q
was only invented a few thousand years ago but once we did
it it turned out to be fairly easy for us to
do and humans
are more or less born babies are born with good
woooo uh and woooo
in general and the problem
is that most woooo is actually uh quite
dangerous for the human mind so
to find good woowoo
so here's one
definition art is
a lie that tells the
and learn
the rules like a pro that's the stuff you can do on the right
hand side of the board so you can break them like artist
like an artist on the leftand side of the
board you got to do both or else
just bullshitting around
and this is hard to do in school because as Marvin
said school is the best thing ever invented to keep you from thinking
about something important for more than a few
minutes you have to ask that about where you are
right now even though I'm a professor
there and Len is a
professor uh schools have their own way about
them like the fact
this talk actually has a
uh designed end back at
MIT they would just get Marvin talking and
out of steam in about six hours or so and everybody
here's another genius huge
influence on me he said institutional
education seeks to make the strange familiar
well that that sounds kind of reasonable
yeah that's what we're you're paying them for
things you don't understand you want to Now understand it but
Jerry says the job of real education
is to make the familiar strange again
it's the stuff that we've that has
come embedded into our belief system about
normal that is keeping us
being pedestrian
thinking would you do if normal could seem strange
normal you would have been able to
escape your present if you escape your present
don't forget to use a
Lifeline how do we measure
things well usually with graphs that
go from the left in the
past towards the future going towards
the right and if something goes up
we say a and if it goes down we say boo and so on and
so forth and if you think about this
graphs like this are complete
and the reason is in order
understand whether we should say yay or not
we have to have a threshold and the threshold
here like these could be uh reading
scores for children and uh
the children are never learning to read here
they're getting better or worse
at what they're not being able to do
and so the key to make
is to understand that both better and perfect are the
enemies of what is actually needed better is the tough one
in engineering school because you get rewards for doing something
better but what we're looking for here is
something new and to get something new we
have to come up with the low the minimal
idea that is just above the
threshold it's just into the
and we've once we've stepped into the blue we're actually in
a different way of thinking about things
so this is you know like going from
uh not having mathematics to mathematics
or not going from not having writing to having
writing or going from not having science
to having science these are big
things so you want to look for that star
and the learning curve you might have to start off
below where other people
think you should be but the main thing is to
up there okay so
here's another example world's greatest
hockey player made many many more goals
than anybody and he was not a big
uh had a couple of nice sayings he missed
100% of the shots you don't take people do
that all the time you don't try enough
and here's this big one a good hockey player goes to where the
puck is a great one goes to where
the puck will be and he didn't mean calculating
the trajectory of the puck he meant getting to a place
where one of his teammates could pass
do a quick shot on goal
that was the Wayne gy game and we can play it in
technology so here's an example from my
own career I had a woowoo intuition
about children having their own
portable tablet computers in
1968 which was impossible
back then but the second
stage is can you find a favorable exponential
Mo's law
and this allows you to take this
intuition out 30 years
and out 30 years in the future
could say you can ask
would it be ridiculous if we didn't have this and
30 years that's 1995 the answer is
yeah it' be ridiculous if we didn't have tablet computers
by 1995
okay so you bring it bring the idea back to
the mid 80s can we do
something by the mid 80s
yes but
software takes a long long time
and we're in 1972 now so
what we need to do is to make a bunch of
flexible supercomputer time machines
personal make a lot of them
so the researchers can live and work in the future
that is what we did at Park that dollar
sign says More's law
means that for most
points in the future you can have
them now if you're willing to spend a lot more
money the commodity thing in the
future is the supercomputer of the present
build the right kind of supercomputer in the present
and you now have years to work on these really hard
software ideas and that is where the star
was at Park
uh the dotted line there says you better
do this really quickly or else you're going to get
submerged into tool building
you're going to be doing uh just computers and operating
systems and you're not going to be doing what you're originally
trying to do so you have to be have to be
have a lot of chops to
do this and in fact the alto was done in a period of
about three and a half months by an AB absolute genius
uh Chuck
ther and once You' got this
you can start inventing the software of the future and
the supercomputer nature of it means
something like user interface
you can you don't
have to optimize you can just
uh do 10 experiments a
day for a week throw away your code
don't even bother stay because uh you could put a higher level
language on it you can do experiments you don't have to
optimize you need to find out what the
how the users respond to things and then
right hand side uh if you're
willing to optimize you can do software like is going to
exis in quantity in the 80s
that happens to be uh
the first version of Microsoft Word which was done at
ZX park it made uh Microsoft
many many billions of dollars when it was
released uh 15 years
later right
okay okay we're coming down to the end here
so 1963 lick wrote a
memo members and Affiliates
of The Intergalactic computer
network they asked them why Intergalactic
he said well Engineers always give you less than you're asking
for so I want a worldwide Network so I'm going to ask
for an Intergalactic one and
when they when they quit bitching and moaning and scale it down it's still going
to be worldwide had
this great line here if we succeed
an Intergalactic Network then our main problem
will be learning to communicate with
aliens we'll look at that for in a
minute and
to this psychologist he said at this extreme
the problem is essentially the one discussed by science
fiction writers how do you get
communication started among totally
uncorrelated Sapient beings and he thought
here of that scaling
more or less destroys communication
even amongst humans and scaling
and technology is going to destroy Communications
and so forth so
this is the this is what we need to think
about and he and Bob Taylor wrote a
paper of which I've expanded on a little
bit so the idea is
uh here's a human
who's trying to say I'm come in
peace and the alien is saying uh
maybe I should kill this
guy between two humans
uh a description of something with a man and ho
Hooves and a tail could be a horse for one
guy and a zebra for
another here's our multiple
BL brains I talked about one part of my
brain wants chocolate the other one realizes I should have an
apple then we've got this new thing
the computer
what is what's on its mind not much
and uh it doesn't understand what a horse is that's
ure so when we're talking
else we might not be able to get together on what a
horse or a zebra is but we have a better
hamburger worldwide so
shared context is what we're looking
for and if we find a way and
what we do with language is to try and point
what we think is the shared context between us
person so if we look back at the theater here
what is it what is
shared what the theater is for is to reflect
the intelligence of the audience back to them you're not
trying to teach them I'm not trying to teach
you right now I'm trying to give you something
that you can reflect your intelligence back
on so the theater has to do
most of the negotiations for the audience that's too
bad but writing and reading are the same
thing same idea
and there's a huge difference between a good writer
and a bad writer for these
reasons so let's think of communication
as an extension of the gesture
so what could two different computers have in common
well could be something like the dynamic messaging
of real objectoriented
programming could be something like what Linda
did but with David gar her we
still have the problem of real aliens but let's look
at the problem with the
person people and computer what
could they have in common
what can we make that is going to beam the
audience's intelligence and knowledge back out
them what should we put on the
and so the guoy is a representation of
what the computer is thinking about
in a form that humans can uh
understand the human can
actually gesture at the
topic that he wants to talk
about each window holds a
context and is a topic for conversation all the possible
conversation are active no modes
undoing is critical everywhere gesturing and
dragging and J Eng gauges the body
brain pictures are more memorable than
words so use those doing with images make
symbols and this is kind of Jerome Bruner
101 this one is particularly
important okay so
this is kind of
a summary
of the way
people were thinking back then
and if you think about the park
guey of all the things that
this research community did back then
the thing that made the trillion ion dollar
Industries was not just Mo's
law because
there was nobody to sell Mo's law or
even any reason to do Mo's law
at the scale it which was done unless
millions and billions of people could actually use this stuff and
reason millions and billions of people can use the stuff
is because of the invention of the
okay so
I don't even know what time it is
but um I'm up to the
dot dot dot question about what my greatest
aha Etc was and but I'm gonna pause
here and just check in with
hello um
how long how long was that by the way uh we
are uh we uh we still have time so
if you have uh
okay so um
the so now now the choice
uh uh can get
a another question from the audience we can
do this dot dot dot thing whatever you'd
like um we have one
question so Len
was this uh was this recognizable to you
oh my god totally I mean there anecdotes all
over the presentation that would add to it
but you nailed it beautifully you captured that
there and the characters
and the uh mix between
highly Cooperative groups and
individual working on their own
things happen there's a lot of individual
yet the group action came in
and out in and out it was an interest it wasn't
pure you know one versus many
it was so
I think I forgot to mention was uh that
I don't think uh lick intended it or maybe
he did but uh what turned out to be the
messenger uh mechanism
and there happened to be graduate students and interns
including you for example
I mean I flew 140,000 miles as a
grad student Dave Evans had a huge budget
for airplane travel for his grad
students well Steve managed
to basically bankrupt mine
doing the same thing yeah because you gota
these and and ARP I should have mentioned
that arpa in uh
1968 funded the first arpa
graduate student conference which was modeled
on the arpa contractors meetings
that's another story about how that happened but
uh that was really great because uh things went
in the amount of intercommunication
and Trust because a lot
about trust and
yet it was not um what shall I
say a
a bowl of soup it was contentious
in a positive way I'll give you one example
we had this principal investigators meeting John McCarthy
arrived about 20 minutes late someone was presenting
minutes later McCarthy raised his hand and says excuse
me I have no idea what you're talking about
he says but I was late
however if I came in in time I still think I with no idea
what you're talking about B can imagine your interaction
was wonderful yeah I should have
that the one of the main things I learned in grad school
and everybody at Park knew was
uh that the purpose of arguing is not to win
but to to illuminate
and uh in fact when people started getting
he in an argument Taylor would
say type two type two
and a type two argument was all
sudden you had to explain the other person's position
back to them until they agreed with
it and then they had to explain your
position back to you until you agreed with it and by
happened you'd forgotten what you're arguing about and
you okay let's go to the question
all right we have a question here yeah hello
um and I I agree and I guess I kind of apologize
for the uh structure
that we have because it does impose any limitations on the
discussion um I'm on on the faculty
too so uh but
you know faculty meetings uh there's too
much contention because the stakes are too
low I
quit going um so I have two questions which are
a little related um so so number
one was
do you think and and maybe you'll agree that
there's kind of been a split between the
practicers of technology which are the software engineers
and the systems thinkers who are like the theoreticians
and the inventors um if so why
any suggestions to remediate this uh
also asking to Professor kleinrock if he has any
input on that and my second question is probably a little
related um which is what are your
thoughts on I guess all of this
development most like a lot of it has been going into private
research Labs like Google X and Microsoft
research um do do you think it's still kind of
analogous to Bell labs in back in the
day uh no
B for instance at Bell Labs
if you walked around back before
the divesture uh you
that would say
either do something very useful or very
I've never seen a sign like that at
research okay so
because that's a pro that's a huge
problem I
guess go ahead yeah go ahead oh no I guess I have a follow-up
question to that which is uh you mentioned a bit earlier in the talk
one of the questions I sent over which is uh do you think
younger people have a harder problem these days in you know living
in this more materialistic world world so
to speak yeah you know
just just one example I I was
at Park in Palo Alto and
the general inflation has been about
a factor of seven since then
or so and when I
factor that into the salary I got at Park it's about what
I would get just you
know I was just a year out of grad school when I went to park
so uh really uh not a
huge wasn't a lot of money but
with today
however I was able to buy a house in
paloalto uh right near
Park and it only cost twice my yearly
okay that house has inflated
by a factor of 70 not seven
so then do you go go ahead
go ahead so the uh
so if Park existed today
so if Park existed today
uh none of us
and and Taylor wanted to hire he only hired older
people uh from the arpa community
as advisers
he didn't want I was the oldest person there and I
was 30 he didn't
he wanted a new know he wanted the researchers that ARA
had Jed out there and he used
the uh the generation before us as as advisers
and so if
if I if they had tried to set this
up uh somebody would have had to
buy a lot of housing in Palo Alo
and you know when when Google is being
set up and when HP was being set up I said hey look
at Stanford Stanford figured this out a long time ago
because if you get an associate professorship at
Stanford one of the perks is a 99-year
lease at a very low interest rate
on Stanford land for a house
and Stanford was owns a lot
houses in Palo Alta they bought them up before the Big
Deck companies got there because they
understood that people have a life might
not be much of a life outside the lab but they still got
one by the way
Lan will love the the story that
metf and bogs had the record
longest period uh
inside of Park without leaving it
which uh during the critical period of the
the ethernet which was two and a half
months they had a
cafeteria you know it had plenty of places you could sleep
it had showers
um wasn't wasn't that helpful on
if you had a marriage that didn't
get this idea uh
they would suffer but yeah so to so
the the a big problem today and and you know
I run run various
kinds of I've been able to fund until
about I don't know five six years ago I was able to fund
a nonprofit research thing that
was about to size of the group I had at Park
and uh and to get people who are
as good as the people I had at Park
but it wasn't the same as Park similarly
Butler Lamson and Chuck ther went with deck
initially and then with
Microsoft and it wasn't the same as park
because they
uh you know for example
my habit when when I would get confused
about something was to go down and have
Butler tear it to shreds
and I miss missmiss
being able to just walk down the hall into
that crystallin Palace of his
mind and you could
never top Butler on the same
least I couldn't but Butler
was an was honest if you came back three days
later and said now here is the real point
you would say you're
right so that's a really great C it's a
genius there's no question he's one of the most brilliant
people I've ever met and an honest
genius and a guy who by the way
helped my group whenever he could
despite the fact that our theory of programming language
design was quite different from
his he didn't give a about
that so he was Butler was a big man
in my opinion just a big fantastic
so it's that Synergy
it's what happens in the
halls and it's
what happens when you're having like
EP uh
you know how how many beer beer Laden lunches
do you have how many bike rides up to the top
of the uh the hills around
Palo Alto with all these people no
there's a whole whole thing there that is going to
pay off by
what the group is willing to do and reason I kept on
showing theater is
arpa at its best park at its best was like
theater at its best you've got the most diverse
group of people you ever saw
some of which some of whom can't even
stand each other
but the whole point of the thing is
audience that's going to be there
six weeks from now or six months
from now and
so you get this thing where people
work 24 hours a day and they just find
ways to solve everything they
do the woo woo on the leftand side that's
what theater is about and they
do all of the stuff you have to do in the physical
world to make it work and
know if you don't have those and
if they don't like Len did
this if those people don't have some of
the other in them to have this shared context
there has to be a place where you go and it's
the love of what you're gonna
get the audience to try to think about
because you can't tell the audience
anything can't people only remember
you make them feel
so so you have to lay it this is why I
all these pictures and stuff because I want the pictures
to do things you can't do in regular
words and
yeah so the question is
uh the way I look
at playing the blame game
here is businesses
are in a terrible position
fidu in a fiduciary responsibility way
for doing the kind of stuff that arpa
and park did they have every
and especially the Milton Freedman American businesses
which are a terrible way to do business in
general there you know Peter Dreer used to say
that that business's main obligation
is to their uh customers
and Milton fredman said no business's
main obligation is to their stockholders
and that has turned really ugly that
is really ugly so but in any case
good or bad they're in a bad place for it
uh research can only be
xpensed eeks
you could be sitting on a ton of cash tough
comes out of the bottom
line so so when I when
I look there I Look to society and
government part of the idea
of getting people to pay
taxes is to deal with things
in a way that is hard for
scattered groups to deal with with or without
engelbart stuff and I think the
the woo woo research stuff done
right is one of those things and I just happened
to stumble into it I didn't go to
Utah knowing anything about arpa I'd never heard of arpa
I just stumbled in there by by and
chance and I found out I was in Paradise
and the dod was doing fine because
uh arpa ipto cost almost nothing compared to
the the whole DOD
budget it was just fingernail clipping so it's the perfect
place for doing this
uh enlightened public
domain uh research and you had the ultimate
enlightened public domain mind in lick
to do that the key thing
Park was that Taylor loved
lick he he was a psychologist
also he truly loved
lick and I'll I'll
finish in one second Jason and he
he studied lick so lick
was intu
intuition he had great inter but Taylor
codified lick those rules
put up there were not Lick's rules but Taylor's
rules and anytime you wanted to park you could
go and lie on the floor in Taylor's office we we were floor culture
we had thick carpets that were
padding underneath so you could just roll around on the floor we had bean
bags and inste of chairs so you go
there and lie in there for an hour and let Taylor tell
you what he was why he was doing what he was doing at
Park and it was mostly these rules
I wrote down you got to do this you got to do that
uh etc etc like
okay so um
yeah so I point I point my finger
at the government and if you point your finger at the government you
have to point your finger at the
uh uh the American
Body politic and if you point your finger at them you have
to point your finger at education
public education if you point your finger at publication
public education you point your finger at the government
again so the government has not
done its job and by the way over here in London in
the UK it's it's worse
uh the politicians are a little nicer over here
but they're equally uh incompetent
they just don't do not understand and
their financial base is much smaller over here
so they're in more desperate need
of the kind of Industries they can afford
that actually work in the 21st century
need this um
yeah so uh but when
after the government
should be doing this that and the other thing
I I can't get far from the educ
from the University because the university used to be the co
the flame the whole
point there was
the universities were not beholden to
business their job was not to do job
training in obsolete software
systems to deal with business's
problems with Legacy systems that is not the job
university that's a trade school and there's nothing
I love trade schools the high school I
went to was a was a one you had to take
a test to get into but it was a glorified trade school Lenny
will tell you and uh I learned a lot
there but Jesus the
universities have got to uphold
something like the way the 21st century
should be for the students or the students
are just going to go there and buy degrees in order
to get jobs that is my general
picture so the university
could put us you know I when I was active
around UCLA at some faculty
meeting uh
maybe more than one
I brought up the idea hey wait a minute UCLA is one of the
top 10 schools and uh
there are other top 10 schools in the nation
in the University of California
you know basically you
know why are we letting the
Regents determine that tenure
uh re requires paper
UCLA should be able to say well you know Screw You
Regents screw you and
Davis and uh San Diego
and Santa Barbara screw you Regents
we're not going to do it that way
we're not gonna count papers for Christ's
sake you know
if you got to count a paper you at least have to put a weight on it
and it's hard to put a weight on a paper until 10 years
have passed you
just can't do it that way or you're going to turn it into
a Job Shop and
uh you know uh I pointed out
to the UCLA faculty that because of
you know the the discussion was wow
our department is really popular now we've got more
students than we can handle I said well godamn it
if you got that that's what the bean counters
in the University care
about is retention because it's money
you've got them just say
okay all right
we're going to do tenure differently we're going to hire
differently we're going to have a different set of Standards
or we're just goingon to uh you know etc
etc but the 60s were
long over by the time I made that
okay next question
I have a question yes
was sorry
this was a and I've heard you many
times and always I I recognize
this accent always
yes I'm ready I'm ready
I'm ready always very entertaining
and uh thoughtful
but I I feel there is now
a huge crra a
Gap uh which we didn't you didn't
would be nice if you could have a thought or address
which is everything in the past
seems to me have to do let's
call it with mathematics has
Precision prob language is not
we talk it has precision
and suddenly I think the new
Queen the new the new thing in the helm
the new Revolution is is
without Precision it's without
and we built machine that we do not
understand how they work but
they work they perform they
perform not in Precision
you know if there will be a a self-driving
car it will kill somebody there is no question you
cannot guarantee that this will not happen
it's a new thing and
I feel most of us here on this list
on the nine people that are left
we are dinosaurs of the past
and what do you have any thoughts about
this well uh
just by chance in the late 60s
I was visiting MIT and there was a meeting
of DOD people with
uh the AI people at MIT
and I think mensky was saying
you know in a few years we might be able to do what a dog can
do and uh the fourstar
general said a dog
he said look we don't even want what ordinary humans can
do we can draft them we've got two million
robots in the in the Department of Defense already
we train them and they learn well and
they carry guns and they do all the stuff that we want
we already got intelligent robots we need
superhuman intelligence that's what
we need and
so I
I don't want to start a long discussion about
you know but it happens that the
what J Judea Pearl calls curve
fitting machine learning
uh as you know
also very well goes all the way back
20 30 40 years
or 50 years or more Minsky's
thesis was actually one of the first perceptrons
so the idea of
of finding complex
transfer functions between lots of inputs and
fewer outputs is has been around for a while
and uh it's definitely a a
big component of what's going on inside of our
brains but
the I I think
the we should ask
more for example uh
I wouldn't want to use an
AI that
couldn't be vetted in
some way and of
course uh the chat Bots
uh can lie whenever it suits them
and by the way I just saw that uh
uh open AI chatbot
has recently gone nuts
which is really fun track this down
it started something went wrong and it's now it's mostly
hallucinating so these are
like human beings these are creators
that's what we are that's what we
got inside there and I believe that
we're we're better than we
were 100,000 years ago mainly
because we invented
methods for thinking
the big one being science
math allows you to in a consistent
way but it
it doesn't have any necessary connection to
reality where science wants to be able
to do uh
crisp thinking in a way that is anchored to
reality such that the uh the
cues might have something to do with reality also
I think that is a very very good thing I think it would help
our politics on both sides of the lanic
tremendously if uh schools started
teaching CH basically doing what engelbart
said we have to insert better methods
than the ones we were born with or
simple uh large language models
have uh in order to
uh be able to use them in
any reasonable way that's my my
got it got it and then we also have one more
question from the audience um I wanted to get a chance to
yeah have him ask hi professor Kate thank you
for your talk today um I'm born in the 21st century
so my uh I guess
thoughts about zorx Park is the only one I know of
Steve Jobs's visit where he commercializes
the guey that you guys invented into the apple one
so I was
yeah so what he saw was what
what he saw was what we did
he visited it was a visit to my
group uh and I was in the
room uh during that visit
and the demo we gave him and
uh we became friends uh partly
as the result of that visit so
ask me any question you want to ask about that
visit uh yeah I mean my first question was going
if you happen to be in the room at all but now I know
uh second of all would be Steve had or
Mr jobs uh had a way of combining design
and pushing the boundaries of engineering through that I
was wondering how Zero's or
the Park's attitude towards pushing research
um the boundaries of research were or what
influence so
the so let me answer it's a two-part
question so one part uh is
early in the 80s I said to both Steve
and Bill Gates uh
this is during the controversy of whether apple and
Microsoft had stolen the ideas from
Xerox and
you know I told I said to both of them I said
look we want this we want you to steal the
ideas but for Christ's sake steal
whole idea don't steal a protest
caricature yeah the whole
yeah that's why I say the the disappointing figures
that I put there was the amount of actual
takeup uh so the second thing is that
what Steve was
uh he was NE say the two words you said again designer
and a pusher of
engineering yeah right okay so
Steve was not a designer
and he was a pusher of
people so the key thing about Steve
was that he had a lot of good
so where he picked up
he could feel the fireflies if he'd been a scientist
it would have been much better for
everybody because as he said and there's a famous
interview you can find on YouTube where
Steve says he says when I first went to Xerox
I was so blinded by the user interface I couldn't see anything
else I couldn't see object-oriented programming I
see the networks he had to
wait until he did next years later when he ran into
another contingent from the old Arbor Community
to explain to them why you why you need
those but we he didn't
understand how the stuff and we couldn't tell
him uh we were
able to show him uh like one of the famous
things that happened in the demo
my my one of my favorite demo things and if I had had time
because we recreated that system just for the shits
of it so it actually runs the system that we showed
him and um I occasionally will'll do
it live in a talk um
but Steve had two complaints while he was and
he was really watching he
had his acolytes around him
uh but Steve Steve was really paying
attention to the thing and he
said you know
I don't like You' never seen a scroll bar before but he
didn't like the fact that as you pulled it it
went line by line he
says can't that go be done
continuously and Dan Engles who was giving the demo
because small talk is live
I mean live down to the operating systems so
Dan was able to do a couple of Clips
here's where the slow scroll bar thing is
uh change a line of code
quar a second later things are scrolling
smoothly and later on
uh when we were showing text
selection um we liked
we tried a lot of different ways of doing it but we' like the
complimenting the characters
which is what basically is done
today uh and Steve didn't like that he said
can't you do an outline
around the you know he's right because it's comp
and white complimenting
is a little bit too in your face
face so he's right there
and so my jaw was on the
floor because Dan Engles went in there
that was not a trivial thing you know because if
you think of the selection
you have to find the actual boundary of the selection and
you have to uh
treat it the right way and Dan went right to the
thing and basically what he did was to
just take the code that comp mented
copy it and shift the
coordinates of the comp of the second
complementation uh diagonally
by two
so you get to so what it would do in it before you could see it would do
compliment then it would shift diagonally do another compliment
and that would leave a a black line around the selection
just what Steve asked for it's like one of
the greatest things I've ever seen done unprepared
in a demo in a program progamming system in my
life that was uh
1979 yeah so that was uh
good and
uh uh Apple didn't steal
you know Apple did do a great thing
as the ba from the basis of that
stuff uh which was the Lisa
done before the mac
and the Lisa was done
uh with Park people who are
uh uh and with
some very smart people who are at Apple like Bill
ainson and the Lisa was
really nice it was the best thing you
could possibly do for that day was
done on a 68,000
microprocessor which meant an enormous amount
of programming for to do the graphics that were so easy
on the alto but they did it and it was really
great and uh a system
you would die for you could get
for 10,000
bucks and $10,000 happen to
be the average price of a car new car in the United
States at that time um sorry to interrupt
we're almost at the top of the second hour uh
of the of the talk so almost at the end of the second hour of
the talk um so I just want to get a quick chance for people who have
class um but first please Round of Applause for our
speaker for the people who are
left yeah
so I
love I I just expected
uh that you would have you would give me
one of your uh Barbed
comments I love
okay thank you thank you than you one and all
thank you for thank you thank
you Alan thank you so much yeah but don't
don't use your car while you're driving Glen
well I had to meet Stella she's waiting for me
and for an hour and now so sorry I had to step
out that's good you listen to
of course happy wife happy
life uh the number one
that's rule zero
not that's not even rule one that's rule
zero you got it you got it
thank you very much again Alan thank you thanks
so much Allan
thank you bye bye