Alan Kay Squeak Presentation (2000)

From Viewpoints Intelligent Archive
Jump to: navigation, search
I shall pay in person but our entire
group is over in Europe attending an object-oriented programming conference
first I thought I would have to cancel but I really wanted to make a
presentation at this meeting so I thought might be a good idea if I
captured a talk that I gave just a few
days before we went off to Europe to the
National Academy of Sciences it's not exactly the same topic as I was going to
talk about at this meeting and after the
talk I will be available by teleconference for questions so thanks a
lot and here we go very happy to be here
I have a theory why the two country
countries that have McDonald's haven't gone to war I'm I think it has to do
with the fat content [Music]
the the name of my profession is computer science but of course it's not
a actually a science it's more of a kind
of an unholy combination of mathematics and engineering the reason it's not
quite a science in the in the usual sense is that quite a bit of science is
analytic that is we're presented with nature and we try and find out how
nature works and one of the ways we've found out how to do it is to try and
take it apart and make models of it and
on the computer you is basically synthetic so we actually have something
like a dynamic mathematics and we create
things by making mathematical models of
them and then the computer carries them out and so it's not exactly clear what
it is that you're when you find out things in computing it's usually either
something that's like mathematics or like engineering or both but in spite of
that I'm actually here to talk to you about science I think Bruce invited me
to give this talk because I've been involved in trying to help children
learn powerful ideas for many many years
and a lot of the work that I've wound up doing with computing has come out of
this interest in children now to give
you an idea of or maybe a motivation for
some of this I like to show you a little clip that NSF the National Science
Foundation took at Harvard University a
self-proclaimed great American
University and I think it is - about eight years
ago I guess it is now so this is at graduation and the seniors had just
flipped their tassels and we're hugging each other and NSF went around with a
camera and asked these poor graduated
seniors a few simple questions about science so let's see what happened to
test how a lifetime of education affects our understanding of science we ask
these recent graduates some simple questions in astronomy consider for
example that the causes of the seasons is a topic taught in every standard
curriculum okay I think the seasons happens because as the earth travels around the Sun it gets nearer to the Sun
which produces warmer weather and gets farther away which producers cover the
weather and that's and hence the seasons how hot it is or how cold it is in any
given time of the year has to do with the the closeness of the earth to the
Sun during the seasonal periods the earth goes round the Sun
and and it gets hotter when we get closer to the Sun and it gets colder
when we get further away from the Sun these graduates like many of us think of
the Earth's orbit as a highly exaggerated ellipse even though the
Earth's orbit is very nearly circular with distance producing virtually no
effect on the seasons we carry with us the strong incorrect belief that
changing distance is responsible for the seasons I took physics planetary motion
and relativity whatsoever and I and I
got through school without having it I've gotten very far without having it I had quite a bit of science in high
school yeah through physics what first year in two years of
chemistry regardless of their science education 21 of the 23 randomly selected
students faculty and alumni of Harvard University revealed misconceptions when
asked to explain either the seasons or the phases of the Moon when it's further
away from the Sun then it gets colder the earth position interferes with the
reflection of the Sun against the moon
so this is a beautiful you can imagine
wonderful we can see that the real curriculum at Harvard must be confidence
101 is that not one of the people
students or faculty who are interviewed ever said I don't know or I'm not sure
or I've heard that and we actually have
to be eternally grateful to all of them for not actually being able to remember
the answer which would have rented this a null experiment because almost
certainly they wouldn't have understood the answer but fortunately they didn't
even remember it so we got this nice nice result and I kept on waiting in
this video for the NSF people to ask what I thought was the next logical
question but they didn't but it happened I had to give a talk over at UCLA a few
weeks after I first saw this and the audience happened to be mostly
graduating seniors and a few first-year graduate soon so I invited them outside
after my talk and to come up one by one and let me interview them on these very
two questions and I was quickly able to of course the question the next obvious
question for the proximity theory is well tell me this when it's summer up
here in the northern hemisphere and you happen to know what the season it is in
the southern hemisphere and everyone did they all knew it was winter and I
believe that the Harvard students were maybe as bright as the UCLA students I'm
sure that they knew this as well and then for the and I got but basically
about the same ratio I got about 95 percent who had serious misconceptions
about one or both of these and for the ones who thought that the earth was
casting a shadow on the moon to get the phases I asked them whether they had
ever seen the the moon and the Sun and the sky at the
same time and almost every one of them had and I asked him whether the moon had
ever been in phase and almost every one
of them said yeah yeah you know occasionally it's in fact it's always in
phase unless they're the opposite ends of the horizon and in both of these
cases there was sometimes a very long pause after they had elicited this
knowledge that they had inside of them to realize that this very knowledge that
they had contradicted the theory that they had just come out with so so I
believe that this is not a science problem that we're seeing here but a
math problem you know in the sense that they had contradictory information
interestingly enough we're not asking for an analogy here either right that's indexed by the same terms
both worse index by seasons of both our
index by phases of the Moon but these pieces of this video that NSF did is
called private universe so that everybody's living inside their own
little universe but even worse it's a it's a property of the way the human
memory and learning works that we tend to compartmentalize knowledge even when
it has indexed the same way so these things are little isolated things and
the the thing that's most resembles our stories and proverbs and the way
proverbs work if you come home from a trip and your significant other is
really glad to see you then what's the
proverb for that absence makes the heart
grow fonder but what if you come home from a trip
and your significant other is not all that happy to see you out of sight out
of mind so this is why most societies
that rely heavily on proverbs are not literate societies because when you
write them down you discover that every proverb has an anti proverb
or there is a proverb for every case and this in fact is what we find in the
Bible that people who use the Bible mm-hmm
almost never read it sequentially but they dip into it for residence and this
is the way we use stories the movie that you loved last week could contradict the movie you
loved today and you don't feel any real problem with it because it's the
business of a story to be a little world in itself and our brains are set up to
learn things in terms of these little nuggets case based reasoning this is why
engineering by the way has been around for a lot longer than science because
you don't have to worry about contradictions to do engineering each
piece of engineering knowledge in the old days before the Scientific
Revolution was a kind of a cookbook and was kept as kind of a cookbook this is
how we learned how to do this and this is how we learned how to do that and so forth so so there's an interesting
problem and I've been interested in this problem for a long time now as mentioned
in the bio back in the 60s I was very
interested in what is it were now called
personal computers this in fact is the first not the first personal computer
that was done by a friend of mine a few years earlier but this is the first one
to actually be called a personal computer and people asked us what we
were doing we used to say well we're making a personal mini computer and that
got too long to be says and this is a picture of it on its own display so you
can see what personal computer looked like back then and we like tablets and
it had multiple windows and it I had
clipping and it had a kind of an iconic interface and so forth so they had a lot
of things that are in the technologies of today and it was partly influenced by
the first pen based system which you may not realize goes back even earlier that
the tablet actually was invented the same year as the mouse in 1964 and in
the ARPA community both of these devices were used quite a bit and so this this
system I ran with one user on a mainframe that
was quite a maybe about four or five times the size of this auditorium and
what we were trying to do is say okay now let's make something for one person
that has this kind of interaction in the summer of 68 I saw this tiny little
one-inch square first flat-screen displayed on the University of Illinois
and we started thinking about gee we
know how many gates are in this desktop machine I wonder when we could put all
of those gates this is about maybe two cubic feet of silicon in this desktop
machine put out a lot of heat had a fan
that was likened to a 747 taking off was
a desktop computer if the desks were made out of steel but when we saw this
we started thinking about gee I wonder when we could put that circuitry on the
back of a display and the answer was at
least 10 years out Moore's law was known about that it was first articulated in
1965 but 10 or more years out I was only
28 and 10 years was and was an infinity so I wasn't interested but a few months
later I visited seymour papert's first work with logo and children and here is
what I saw here's a little program that
papper could get people to do and did it
by getting the even kids as young as
five years old to do the following thing close your eyes make a circle with your
body and when the kid was about halfway around Seymour would ask the kid what
are you doing and the kid would say oh I'm going a little and turning a little
over and over again and so in logo two
circle means repeat which means do it over and over again go forward a little
turn riedel over and over again and what you get is
this nice little thing and my background
is mentioned was in mathematics so I looked at this thing and I thought wow
now of course I knew all about this but what pepper was showing us was that a
kid's toy that was actually doing differential geometry not even crass
cartesian stuff but this turtle is an actual vector in a different in a
differential geometry space the actual scientific language of much of physics
and I just went bananas I thought this
is the best idea anybody ever have for what computers should be turned into
which is an actual Montessori
environment so on the on the way back I
on the plane I started thinking about a computer for kids and made this
cardboard model in 1968 because it couldn't be a desktop computer for kids
would have to be something the kids could take around it would have one of these flatscreen displays on it and it
would be something that had authoring as symmetric as consumption so it'd be a
thing that kids could learn things by making them and it would be a Montessori
kind of environment a Montessori idea was that children in the in the natural
state in traditional societies are set up to learn the place they're in through
making up games imitating adult behavior through games
and she said if we want them to grow up she said in the 20th century if one of
them to grow up in the 20th century instead of the eleventh we have to give them 20th century or 21st century places
to learn by playing and so this is what caught my imagination that there's a
possibility of making a medium and not
just a medium a meta medium one in which Nuuk descriptions generate new
interesting thing that the children could learn powerful
ideas on it happened that very summer we're also about maybe 60 of us in the
ARPA community we're also starting to design the ARPANET which by the way
started working the next year so the the Internet is actually a little over 30
years old now had its 30th anniversary
last year in September and we realized
that we were going to have something like the printing press that could spread ideas all over the world in this
new dynamic form and so we had the sense
of a progression from these really expensive manuscripts of costing
millions of dollars in today's money to Gutenberg Bibles which tried to imitate
them as much as possible but only cost sixty thousand dollars in today's money
you might ask how does he know that Gutenberg Bibles cost sixty thousand
dollars well at the Nuremberg Book Fair they marveled about two things one was
how alike the twenty four Gutenberg Bibles that were displayed there were of
course they were exactly alike except for the illuminations which were done by
hand so that people wouldn't be afraid of them they wanted them to imitate
these old ways and the other thing was that these things only cost three years
of a clerks wages which is about sixty
thousand dollars which is nothing so you can think of this as kind of a maybe a kind of a mainframe a couple of million
bucks and only the Catholic Church or somebody really rich like the brother of
the King of France could own it this is kind of a work station then eldest came
along about 50 years after the printing press with these books that are the size they are today I wonder if anybody in
the audience here knows why books are this size today is sort of a standard
book size it's a really wonderful story that Aldous was a scholar who decided to
set up a press in Venice and his his I
meet his new idea was what he called the portable library so he decided the first
40 titles that his press printed were
called the portable library and they were not religious texts and the reason
the books are the size they are today is that in 1495 Aldous went out unto the
streets of Venice and measured people's saddlebags to see what was the smallest
saddlebag that was kind of common and he decided he would make his books fit in
that now when you see a book you'll never forget why they're that size so
Aldous hay and Elvis's books only cost a few hundred dollars in today's money so
they had a much smaller types size he had to invent new fonts that were
readable and so forth but then the interesting thing to me about the
printing press was that the big payoff on the press lagged by a hundred two
hundred and fifty years the big change
from it came in a change in how people represented ideas and argued about them
this happened in the 17th century and then politically in the 18th century and
so we had the sense that if the computer networked through what's now the
internet was going to have a significant change to humans that we'd all be dead
by the time this interesting thing happened because our belief was that
people would just try and basically do the Gutenberg Bible over and over again
it was to imitate old forms for a long
long time and so the question was can we actually speed this up and this is what
we tried to do at Xerox PARC was to actually try and guess what the media of
the 21st century would be like and to do it early and it's led to a lot of
technology most of which I think is recognizable the forerunner of both
workstations in the Macintosh from 1973
the laser printer object-oriented stuff
the Windows interface many of the software the Ethernet was done there and
pierre-pierre and client-server architectures and so we're kind of
aiming somewhere between this desktop way of looking at things and this new
laptop and I turned out there was a wind up
being an awful lot of work just to get a few computers into schools because that was what most of this technology was
designed for it was an effort to get
something that the screen had the shape of this eventual notebook computer that
we could make enough of them we eventually made about 2000 of them in the in the 70s so that we could get them
out to people who are not computer people and start learning about media
okay and so this is kind of a metaphor for what we had and had in mind there
and I'll show you a some legacy from the
past and also some modern versions so you might notice that this is actually a
Macintosh hardware I'm using but no
piece of Apple software is running on it right now or has been through the entire presentation so what I'm actually using
is a what you might call a 21st century
version of the software we did at Xerox PARC so they'll give you a kind of a
glimpse and the main difference between what I'm showing you now and then is
that there's now enough bits to be able to do color so we only add black and
white back then but now we have color but pretty much like for instance this
is the here's a Scientific American article I wrote about this subject it's
written in the very first desktop publishing system going back in the 70s
you notice it's pretty smart can feel the text that I'm putting stuff in and
here's a little flap of tools you know I'll drop paint palette in there and it
gives me an onion skin in case I want to trace something out I'll just do a corporate logo
and as soon as the thing I paint it
actually becomes an object that has a shape that's recognizable by the system
and it can feel it can feel it and the
system has a kind of a standard way of
being used here so that when I drag through texts it understands that I want
to select something but because this is a system also for kids we've made it D
constructible so there's a fence that
keeps the kids from getting hurt but they can climb the fence whenever they
want and these dynamic relationships are actually maintained so if I scroll this
guy you can see it's scrolling through that thing that I took apart okay so
that give you a kind of an idea there
and we'll look at a couple of other things just to give you an idea one of
the nice things about being at Disney is you they're all these artists around so
this is an example of two-dimensional media and there's some special things
about it for instance here's a bitmap version of this poo which in 32-bit
graphics is actually 41 K so he's actually kind of big and he has a couple
of bad things happen he goes up to about 500 K here and I make him big and it
gets Jaggi so both of those are bad in fact bitmap graphics are not really what
you want when you're dealing with the Internet let's take this guy out of here
the identical poo and let's ask him how
big he is and he says oh I'm only a K long okay big and if I blow him up then
something interesting happens is that he always stays at maximum resolution and
if I ask him how big are you now he says
I want to still only a K by long so this is this is an audience that can actually
understand this this is like writing a circle is all the points equidistant
from a given point on two sheets of paper and giving them to two people and
saying make one that's six inches in diameter make one six miles in diameter
in the description say is the same size so this poo was drawn by an artist but
it's captured and render it into a resolution independent mathematical form
and you can sort of just to give you a
sense here I'll pull out my little magnifying glass here and you can
compare the two edges you can see that this this edge here is feathered as well
but it's feathered in a way that's always dependent on the on the actual
pixel size that's there remember I said
this is a meta medium and everything that's in a computer is created out of
these new what you might call new mathematic description so there's also a 3d system
in here and all of these systems are scriptable so for instance I can come
down here and say the bunny move forward good an undo here turn your head around
once here's a crowd-pleaser which is
always look at the mouse and the 2d
world and the 3d world are mixed in this system so you think of what we have here
is a desktop but member the bunny is in 3d and the desktop is usually in 2d and
this system you can this shows you kind of what the relationship actually is
that the desktop is kind of a place in a in a larger 3d world now let's look at
the age kids that we spend most of our time with we work with thousands of kids
since about 1973 and partly for reasons
that you saw in the Harvard video we found that basically trying to up in a
sense update Montessori with
environments that involve science and math toys is a very very good idea
because Montessori point which Piaget later took up is that young kids are
actually not in great shape for learning adult math but they're in great shape
for being taught uncommon sense so they
pick up a set of heuristics and points
of view and ways of looking at the world and ways of thinking about knowledge
from their travels in the world and these are what are so disastrous for
these Harvard kids and most kids and if
you trace it back there's a famous book called science learning by Roger Freitag
as a New Zealander that the attitude
towards how the world works is firmly set in most kids by around the age
age of seven or so so over the years we've gotten very interested in working
with young kids and we're interested in
how why do kids build things small
percentage of kids get interested in ideas and build things because they're
intrinsically interesting and these kids are what you might call experimental error for pedagogical experiments so if
you teach programming in a classroom or almost anything in a classroom about
five percent of the kids will be somewhat pre-wired if you do sports in a
classroom about five percent of the kids will be pre-wired usually a different set of kids and for first couple of
years at Xerox PARC we used to only
point at these kits look at what this kid is doing it's incredible so a couple
years later we discovered that our pedagogy was actually meaningless
because these kids would have learned if we were twice as bad as we were they
just got a fact they were like us everybody in my group had understood
what a computer was and had learned how to program in about a week because it
was just there it is so it couldn't be more obvious so the interesting thing
about pedagogy for any of these areas like math and science and programming is
what do you do with the next to 80% the 80% who aren't pre-wired for the stuff
but can learn the precursors of this stuff and then be able to do it so this
is where we discovered our pedagogy was pretty bad so here's here's my nine-year-old nephew constructing a very
complex thing from standard Lego blocks which is actually a walking dragon so
it's standard Lego blocks articulated and it has motors it has a couple of
worm gears in here it took this kid 55 minutes of deep concentration to build
this and he had four times when he ran out of short-term memory and got
confused he had actually not made a mistake but he thought he had I'm sure
you've had this experience when dealing with children and so the coaching job
was just to get him to slow down Woori go back reconfirm that everything
is going alright and a lot of the reason this kid who is not particularly
intrinsically interested in ideas does things is because he's trying to prove
to himself and everybody else that he's in the world he wants to be an effector
he wants to be enfranchised like adults
so he always wants to do the driving and he likes to drive when he's on on a
computer so here's an interesting little activity which overlaps with what I just
showed you from Seymour for this kind of kid who's not particularly interested in
turtle graphics but loves the idea of designing his own car again we have this
little painting system yeah do the car of your choice as seen from the top and
since I've come to Disney I've learned to do specular reflections because
everything should Sparkle and
interesting both boys and girls these days invariably put big off-road type
wheels on their cars this is a very interesting thing because they're
interested in power so again this makes
the system makes this into an object and
I can do things to the object and here's the game we're playing this is the area
this hand eye areas where the kid gets their payoff and what we're trying to do
without telling them we're teaching the math is to get them to understand the
power of symbols so we have two
identical representations here to make this a little bit more clear I'll call
rename this guy car and if you look here
cars heading is 0 it means it's pointing straight up if I take this little blue
handle and rotate it you can see it going positive and negative in fact if I
go over here to the carrot and start pressing on it the car will rotate so I
have two identical use but one is symbolic and one isn't so
the symbolic one is not very interesting to the kids even when you can press on
the explanation point here and get it to go forward by five or turn by five still
not to any interesting but when they find out that you can just drag out
little pieces of program here and start
it ticking and the car starts going and
in fact you can here's a nice little
motor sound more like if of course
you're not driving it yet so we should do an experiment here first paid
and ever the steering wheel is the same kind of thing as the car is I call it
steer here and it has a headache and it
goes Pleasant - when I turn it so let's see what happens to the car when I say
turn by zero well it starts going
straight and when I say turn negative it starts moving the other direction so it
seems like if I could get the numbers from the steering wheel into this car
turned by I can do that by grabbing this and just dropping it in then the car
should feel the steering wheel and I should be able to drive the car by turning the steering wheel the kids love
this and then they start complaining they say wow this is hard to control we
say yeah it's because you made a race car you don't have any ears between the steering wheel and the car and they say
well what's it here you say we'll see this little triangle here once you press on it you can do some arithmetic
arithmetic ah they say but we say just
divide that number coming out of the steering wheel by three oh yeah it's
much easier to control because now I have to turn the steering wheel three
times this far
those little activity goes on and on and then I could do all of it for you but
just to give you an idea and you'll recognize the seymour papert turtle here
but now with a costume that is meaningful to most
children and the next thing they do like
you can see over here as they draw a road I'm gonna draw a road in a in a
second also the cars can sense the
colors underneath them so they can make rude noises when you drive off the road and so the kids spend about ten minutes
doing this and I have about an hour of
playtime or something when you ask them what was the most fun thing about this
project they say learning to drive the car if you ask them what was the most
difficult thing about this project they say learning to drive the car so but
what they are getting a sense of is the power of having names for things and
also for the first time at least in America for nine and ten-year-olds while
you might ever want to use divide because if you think about it in a kid's
world they don't need divided for dividing up M&Ms because you just do
them sequentially they don't need divided for dividing up a pizza because you do that by eye so what does a nine
or ten-year-old actually use divided for and the answer is nothing and most kids
go all the way through high school without ever realizing that multiplication and division are for
scaling things and that there are many important uses of this okay so let me
let me show you it instead of taking that path let me show you another
powerful idea here I'll draw draw a
little road
okay and now what I'm interested in is
making this car a little smarter so this is usually at the end of several weeks
of experience with this thing so what I'm going to do is I used to sail boats
so the way you remember a port and starboard is that port and left in
English have the same number of letters so port is always on the left hand side
and it's red because of port wine is red
so put into my little red a little red
guy as a sensor and then starboard is
all the other things that's green and on the on the other side okay so here's my
car now it has has some sensors and now
we have to think about how we would
program this first thing we want to do is get rid of this car turn by the
steering wheel so we'll put this in the thing there and the car is going to go
forward where we're gonna make the motor sound so we'll just leave this script alone now what I'm going to show you is
something I think is very interesting the standard way a computer scientist
would go through the next phase is to do a nested conditional so they'd say if my
red sensor is touching something rather than do something else if my green
sensor is touching is a typical thing this is the way we wrote it this is the
way the teacher wrote it now it happens that this teacher is a wonderful teacher
she doesn't show the kids her solutions so she showed them a car successfully
doing working the road with two sensors but she didn't show them her code she
said this is a challenge so this is one of the this is the how the monkey found
the fifth way out of the room it turns
out that the kids these age really can't do nested conditionals it's too
complicates it's hard for a lot of us to react so remember these things so here's
what a number of kids did on this and I
think you'll find this interest I hope so so let's get a new script here
and we'll put a test in it and what
we're going to do here is we're going to
pick a color on the car so we pick we're
going to pick the red the actual red that I used here that one and here we're
going to pick so we're going to say when red sees brown what do we want to do we
can see that the car here if the red sees brown then we want to turn away
from it and the kids use a clockwise
system here so we're going to turn by
minus five okay and then one more little
script here and same basic idea color
C's get the conditional first by the way
this this way of letting the kids do these things without having to type is
actually very important because it allows them to do many many more
experiments so in this case we're going to pick up the green color I used here
that one and again so when the green guy
sees Brown this Brown what do we want to
do well in this case if we see it then we want to turn away from it so let's do
car turn by five let's slow it down down a bit so we can see what to do with it
like that get this guy to be paused and
get this guy to be paused and let's let her go
very quickly some of the kids decided wow you should add some sounds in here
cuz then we can say so let's pick will
do clink for read novel
let's do scratch okay so now let's go
okay now what happens here is some of
the kids realize hey wait this is what ants do and they follow trails and this is what salmon do and they're fine
so immediately some of the kids go and redraw the car which they can do by just
going to that and turn the car into a
fish or turn the car into an animal and so one branch of this is to go into
biological systems so shown you three of
four of the powerful ideas we work with I've shown you vectors shown you
basically differential equations in their computer form and the integration
of them I've shown you feedback systems which is how how you do things without
now notice that elegant this code is by
the kids know what computer scientist is just not used to writing parallel code
so we would never think of doing it this way but when I saw this program from the
kids I just about wept it was just beautiful because they realize that
these cases were independent and therefore they could just let all three
scripts run without needing any synchronization except these tests and
the rhapsodies of a computer scientist
let me show you the fourth thing that we
do here I'll skip past this
this is actually a system that older kids construct it's again lots of little
cars running around and bouncing off things and you can do for older kids you
can do gas laws and other kinds of things and here we're going to do
something else we're going to infect one of them the red guy and let him infect
other everybody always routes for the
last one there he is and to help us
understand what happened we get a curve like that and that just happens to be the characteristic curve of an AIDS
infection it has this long flat period
where nothing much happens and to contrast that we can heat up this
particle system here by just making it smaller and we'll do it again and we get
this which is a lot more like typhoid and cholera and I think everybody in
this room realizes that more than 30 million people in Africa alone are
infected with this disease needlessly
and this is because the unaided human common sense is a keyhole that's only
that big and so if you look at AIDS when
it's starting up with simple common sense nothing looks bad whereas if you
look at typhoid or cholera is starting up you can tell you should probably do something and the whole point I believe
of learning the scientific outlook besides the fun of science itself is
this larger window for imagination it allows us to imagine disasters early
enough to take action and that's what happened in this in this country ok so
now the last part of this I just want to show you following one of these powerful
ideas which is called increase by
start with very young kids like six years old this kid has drawn a grasshopper here
and has made some grass and the the
grass here we can copy
we can make a lot of copies of it and
hide this little grasshopper now six year olds it's an interesting thing
about what can what can six year olds actually do here and see if I can find
the grass up or here come on here it is
but a very wonderful little game can be made with increased by so I've take the
hoppers right left coordinate I can determine that that's the right left
coordinate by getting a little watcher
for it and I can try moving this grasshopper around and see what happens
to it so going this way it gets in it increases so I cover it up again with
with grass and then I write a little script here let's do hopper increased by
one and now I can bring my mother over and say what's hiding in the grass mom
and the mom says I don't know and the kid says hey watch this I think
biologists here for sure know that there's a huge difference between the
motion sensors in our visual system and our ability to recognize patterns we
have special wiring for dealing with these things and this is something that
now becomes part of this very young kids common sense what about animation we get
the kid to draw a straight arrow here
draw a curved arrow and what we want to
do is alternate back and forth between this and have this guy where where the
costume so how do we do that
okay what we do is again right when a
little program and now what we're going to do is increase not how this thing is
moving we're going to increase where this cursor is in the structures are
going to say cursor increased by one and when I'm step that alternates back and
forth and now all I have to do is to
come in here and this actor on the stage
is called Harrison I don't want him to look like a dot I want him to look like
the value that's under the cursor and
now if I say step but if I tell it to go
I get a surprise it actually goes too fast so one of the things I can do here
is use this little green copy guy to
copy more in now I get a nice animation
and I can see that I'm spending three-fifths of my time on the straight
arrow and to end two fifths of my time on the curve there and that's giving me a nice animation so this is kind of
standard animation but here's another experiment I can do what happens if I
change this number here to 1/2
and what is actually happening there well let's let's take a look this is a little complicated here so I should get
a watcher here oh yeah so it's four then
four point five and five and five point five then one and then one so it's
dwelling for two ticks on each one of
these things and and therefore it is traversing through there at half speed
so this is teaching the kids sampling
same program works with a multiple thing so here's this bird and learn to draw a
little bit better pause this guy so
he'll just stick here and again we can
do these experiments we can try let's go up to two here so now he's just
eliminating these guys because he's hopping over them or I can come back to
my 0.5 experiment I can go in my
supplies flap here and I can get out of slider
and if we look at the slider we see that
its minimum value is zero and its maximum value is one its current new
valent numeric value of zero point three so whenever I just pick up the slider
and just plug this in here and now I've
got a continuous control over the speed of this thing okay so this one I'm just
going to point to just the point this is exactly the same as the animation except
we have a timer that we have recorded and when I say go through it it's going
through the air and then the cone position of the speaker is being moved back and forth again I can speed it up
by a factor of 2 and a factor of 3 it goes faster but I can't hear anything
that's the reason is is that I have to actually speed it up here by a factor of
10,000 to go through and if I go up to a
factor of 2 okay so that kids take these
things make up the magic numbers and now now they have the magic numbers to
sample out the different notes put each one on a button and now they have a
little synthesizer
and there's a trombone in there it gets
louder when I bring in more if I don't
want to try but what if I want this little time where I just made so every
sound the kid makes is inserted into
what you might think of as the adult media here except this is all done for
children okay so I think it's a good
place to quit what I've been the next
stage in this increase by is to do
gravity experiments and the way we do science is by actual reference to the
real world I don't believe you can learn science from a book because you you can
write lies in a book I don't believe you
can learn science on a computer because the computer can simulate an inverse cube law just as easily as it can
simulate an inverse square law so you're always basically put in the position of
having to believe somebody in mathematics Einstein had this nice
saying he said you have to be able to learn how to distinguish between what is
true and what is real and what he meant is that what is true is what you can put
in language in mathematics that's because the truth comes from its non
contradiction stuff but what is true is
not necessarily what is real what is real is the way the universe wants to be
and they're really two different things and so we do it by actual references to
the physical world like somebody dropping water balloon and then having
the kid
measure using this little measure so for
instance if you measure from the middle of the balloon to the middle of the next balloon this is every 10 frames here
they quickly see that they have to increase this measure by some amount
each time
and then by stacking these guys up they say oh yeah the amount that's getting
increased is just about constant and
that leads them to this two-stage increase by which gives you gravity and
in the music world a two-stage increase by gives you FM synthesis which is the
synthesis that's used by the Yamaha synthesizer so the belief here and this
is justified by quite a bit of experience that simply putting kids into
a play world that has powerful things
powerful ideas that are used over and over again in different contexts that
these become part of the kids heuristic thinking and I believe that if we're
ever going to have more than a few percentage of the population really
understand science that that's what we
have to do we have to change kids common sense into uncommon sense thank you
Thanks so I should mention that this software I'm doing is free this is not a
Disney product Disney is sponsoring this research but it is actually an
open-source system like Linux done by a
small core group of seven of us at Disney and well if you count all the
people who find bugs for us probably several thousand people out in the
network all over the world in two or three hundred people who put in several
hours every day helping on this we've been working this runs on every platform
runs on little PDAs runs on phones runs
on the Sony Playstation 2 and it runs
exactly identically on each platform unlike we're not competing with Java but
if you've used Java you may be aware that doesn't necessarily run the same on
one platform to another so this uses some software techniques developed at
Xerox PARC to make it absolutely compatible and it doesn't use any native
tools so the graphics is exactly the same the 3d was being done by us all of
the sound was being done by us so it doesn't need sound cards or anything
else so it's kind of a universal software system that's freely available
over the internet and this media that I've been showing you every single thing
I was showing you is something that you would in essentially running out of a browser because once you find one of
these things in a browser it takes over the machine in a benign way expands to
fill the screen and it becomes a new world for you to do work in you can
escape from this world back to your the arms of the operating system you so love
but basically this attempts to solve a
whole bunch of problems of putting media out on the net and allowing it to be
used by anybody authoring is always on so when you're using this thing
regardless of whether you got it through a browser all the things that I showed you are working continuously you can
think of as an attempt to restore the symmetry between authoring
and consumption that we had in personal computing 25 years ago but got taken
away by the web the web has this horrible thing that the Brout web
browsers don't really want you to be an author and so you tend basically most of
the millions of people on the Internet are basically consumers not authors and we think this is just really bad goes
against everything that we tried to create 30 years ago so any questions
okay the the hackers version so I should
mention that this thing that I showed you is squeak org Suite org squ
EA k is the home site for this open
source project and all of the different
platforms are supported there this demo
was done by taking what you can download there and adding only about a hundred
lines of code to it so everything I
showed you is actually in there but none of the media is right now so for
instance the ability to make that car is in there but the project that's set up
to help you make the car is not in there but it will be by the end of the summer
so the reason I'm going around and showing this now particularly to
scientists is because you happen to be one of the major groups that we made the
system for it's for kids ultimately but we actually need thousand great pieces
of content like the drive a car and the maker robot car and they make your own
tomber and the stuff like this and we made this system is a kind of a super
hyper card if you will as an as an
authoring tool both for children but also for people who understand this
content to do free content there's
nothing in the license that prevents you from making money on this and some
people will some people will make money but this is basically an internet kind
of thing where the thing that children most need in this
country at least is something like the San Francisco Exploratorium which is 500
ways to enter into a single idea because
every child is different most children aren't interested in the ideas directly they're interested in
some part of their own personal fantasies and so when Frank Oppenheimer
made the Exploratorium he wanted to teach the children that the world is not
as it seems and every child was
different so he put 500 exhibits in the Exploratorium all of which teach the
world is not as it seems and he figured that if he had 2,000 kids careening
around the Exploratorium that the Brownian motion of these kids against
the 500 exhibits would often collide the right kid with the right exhibit and
they would have an experience now I'll tell you justice here's a here's how deep this
goes this feedback experience is a big
one for the kids just the idea that you can be strategic is a huge thing for a
nine-year-old child that you don't have to know the exact things you can draw
any path and their little robot car will follow it but here's a biggie
maybe even a bigger one because it's so tiny is that a girl was experimenting
with forward so she was doing forward over and over again forward five going
real fast forward - it was slower she clicked it down - forward zero stop she
went to the negatives and went backwards she went up - forward zero and then she
had this epiphany and that was that the car was still moving when it was at
forward zero that that zero was actually a valid way to measure a length that it
was actually something and it's just one of the great things when a kid has this
thing it's not nothing it's something
that's the reaction and so just in these
tiny ways every time you open one of those cracks a great thing happens
because there's a chance for more stuff to come in and I believe that's an essay at least
in this country what we need is a thousand or more pieces of cotton we
have people like Richard Dawkins who are going to do some things and Douglas Adams who wrote Hitchhiker's Guide to
the galaxy you're going to do some things in this is really for the the
scientists and knowledgeable people of the world who are interested in kids to
make a little project that will help
some kids understand a powerful idea so
that's the plan yeah why I don't know
what you mean by the same playground there well we're the it recapitulates
everything you can do on a personal computer that I'd say that's large right
I mean the computer has a lot of degrees of freedom and nothing is restricting
here this is not a stripped-down system that's the whole point of it you have to
have low threshold but you must not have any ceiling so the system is more has in
many many directions as much more capacity than the software you have on
your PC or Mac and it should because it was done it was the original software
for this stuff it was done with a much larger view than to provide stuff for
businessmen or people who want to imitate paper so I would say that it
it's a playground as large as what books are except its dynamic which is another
dimension to it and I don't believe that children are harmed by reading books just because they have characters on
paper every book is somewhat similar you must admit
other questions well it's it varies
because children certainly up through
the age of 11 or 12 often are as many as
two years apart developmentally for the same chronological age you have to be
really careful with this one of the schools that we work with groups kids in
their classrooms not by chronological age the way it's usually done in the US
but by maturity so they take the kids that are roughly at the same place in
development and have them in the classrooms and that makes it a lot
easier but if you're in a standard classroom you really don't want to put a
uniform thing on all the kids in the classroom because some of the kids just
aren't set up for it yet and I believe this is a big problem in a lot of
standard curriculum the other thing is is that the kinesthetic return on
learning for young kids is tremendously important and we've gone so far as to
make for instance mice that will let you
feel the world that you're in but
generally speaking young kids should be spending most of their time doing
physical hand-eye things because these are the precursors for an enormous
number of things it should should be spending a lot of time drawing with ordinary media because even with a very
expensive setup it's difficult for the kids to draw as well as they do notice
this thing we we don't emphasize the drawing very much and it's frustrating
for kids when they try and do a drawing masterpiece even in really great
painting systems and we actually invented the first painting system in
Sirach spark for kids to drawing so we've had a lot of experience at it so
but on the other hand some kids are ready to read when they're three some
kids are ready to read when they're five some kids are ready to read when they're seven and generally speaking when a kid
is ready to read they're ready for this stuff as well and then the thing is just to be gentle
so I have math books in the United States are the most disgusting things I
think if you're a mathematician because they are designed to sell to non math
people who are impressed by having zillions of things in many colors on
each page for hundreds of pages and also
for a variety of reasons the there's an attempt to reteach almost everything
every grade which is very very tough to
pull off whereas my belief in math especially when you're starting out what
you need to do is to understand a few important things about the power of it and the reach of it and I also happen to
believe it doesn't much matter what your first math experience is as long as you
have one suit when you bridge that it's
like what is your what does it matter what is your first book that you read
thing is that there's a before you read your first book you're one kind of
person and after you read your first book you're another kind of person and that's the most important change there
is and so that's why we need to have thousand like with Frank and I we need
to have many many entry points and like a good liberal arts education it doesn't
shouldn't matter exactly where you're starting because if you go deep enough
from wherever you're starting you'll start getting to the good stuff
human beings have not thought up that many great ideas over the last two
thousand years so that's which is good because you have
a chance to sample a lot of them that real trick is getting into them yes well
I mean wait the it's interesting at at
Disney of course is a is an entertainment company and I went there
because of Imagineering which is a the
group of wonderful interesting people who've designed all the theme parks I
was very I'm very interested in the
theatrics of place I think the internet
needs more of the theatrics of place but the way I explained this to to Michael
Eisner is that there's soft fun and hard
fun soft fun is watching somebody play
baseball on television hard fun is learning to hit a baseball
soft fund is listening to pop music on a record hard fund is learning to play
screaming rock guitar or classical music
and then there's the interesting thing about hard fun is that it's intense so
you can't do it 24 hours a day so there's a place for both hard fun and
soft fun and everybody's lives and my complaint about the late 20th century is
that the balance between safa and hard fun is way off way way off because
almost every technology that's come along has provided another way of
providing soft fun for people and I have no objection to every single one of
those but I have a Norma's objection to the amount of time that's spent doing
that so one of the easy ways of
predicting the different pathways where
things like movies theatrical experiences like movies go is just to
take the three or four dimensions that are important in all of these which have
to do with the kind of fantasies that people want to make the kind of work
they're willing to do in a theatrical experience to make it work for them
versus the amount of work that is done for them and the level of aspirations of the
artists that are making the things and I think everybody would agree that
regardless of what you do regardless of whether you allow other senses in the
thing of different kinds of interactions that the most important thing about art
is that it shouldn't be too explicit because the more explicit it is the more
it robs the person of real participation so movies have a real problem today
because they try and show you explicit scenes about everything and so it's
always a shock to read a book and realize how involving it is even though
there's just these boring little black smudges on the page right and so so as
Leonard Bernstein said about classical music he wrote a book called the unanswered question and he just pointed
out that a popular song tries to answer a question it's like a proverb and
classical music tries to leave the question open so the themes go and go
and go and then they pull back or something else comes in and something else comes in because classical music is
more a lot more like the complexities of life there's more dimensions and
textures to it and it's a somewhat of an acquired taste as opposed to say going
to church and having the minister tell you everything is going to be alright because God is protecting us right so
that's that's another form of theater
that is very comforting to people but to
me it doesn't have the actual involvement in grappling with real problems that real life does yeah
[Music] yeah yes we have we've done it
periodically the the hardest thing and doing this
experiment these experiments I said we've done Oh four or five waves of
major you have to do them longitudinally to get anything out of it so you have to
be able to do them over a couple of years at least we have done several
three-year experiment experiments we've done some seven-year experiments you
have to see a lot of kids the kids have to come from some sort of background
that doesn't have artifacts in it and so
we we wrote a little bit about I wrote a little bit better in the Scientific American article September 1991 but the
the nice thing is that that we've always gone for sponsors like prints we've
never been sponsored by NSF because NSF likes you to tell them what you're going
to do in the proposal and they also like
results and the thing is if you're doing honest experiments in this area you just
generally don't get good results either positive or negative you're lucky if you
get an idea and so the people who have been and we're not the only group that
has taken this particular route of just
saying hey we don't care we're not doing this for the sponsors we'll get the sponsors to sponsor us for
other reasons like inventing new computer media and but this is our
application here is to try and find out more about the way kids learn and what
is actually going on and so once you put that filter on it the amount of
knowledge you get out of incredible amounts of work because you can imagine putting an experimental software into a
classroom of 60 kids what do you have to
do to help the teacher how good does the software have to be if 60 kids are
actually poking at it when you're not looking it's like milspec
and yet this is software that you're going to tear up and throw away in a
couple of weeks because you've learned that something doesn't work so it takes
a very special group of people who are willing to just write code and write
code and write code and throw it away and throw it away and throw it away and
but gradually over the years we've learned more and more this experiment
that we just did this year in a school and you know Los Angeles was the best
experiment we've done and not in small part because I didn't go to the
classroom it was actually done with a
system that we didn't think was ready for the kids yet but the teacher wanted to do it we tried to talk her out of it
so I sort of said well we and we can't do this we can't do that and then my
cognitive scientist Kim Rose I decided
she would support the teacher and they found a rhythm that wasn't there when I
was in the classroom I realized sort of I know this intellectually but I have a
problem with wanting to get to it and
what happened was they found a slower rhythm than the one that I usually do
and it just fit everybody it fit the kids that fit the teacher the teacher
was not anxious about doing stuff she was able to be one of the participants
in this thing and I met with them for
about three hours about every three weeks or so just to hear about it and
stuff like that and the next time I do this I'm not I'm gonna do the same thing
because this thing we got some great results and we're going to put all of
these results out including all of the kids actual projects that they did and
you'll be able to look at them this summer and just keep on checking in at
squeak org some sometime around in August we're just going to put two out
this stuff and you those projects will not come to you it just as pictures
they'll come to you as the actual projects that the kids did you can
interact with them you can explore what the kids did we got testimonials from
the kids and all this stuff and that's the kind of results you actually get we
have tried transfer experiments
so well transfer experiment might and this is yeah that's a good question what
does it mean so might mean well let's
give to a year of this stuff for kids and see if they can learn molecular
biology and middle school more easily let's pick something it has a similar
structure like one of the things we did
to lead into this for many years was to give the the third graders a year long
thing in city buildings so they're learning about building complex
structures in the physical world scaling them they would blow them up and wear
them as costumes and a whole bunch of things and I think that does work that
you can point to some transfer but if you think about what's the what kind of
transfer you're getting you have this very interesting thing that Bruner pointed out Jerry Berner pointed out 40
years ago the most interesting thing about transfer is that most kids can't
see the analogies and don't see the analogies until you say there is an
analogy and then it's amazing they just
don't see it it's right in front of their face soon as you say there's an
analogy between this and what you did yesterday then they will draw the things
they don't have any heuristics like they don't have a heuristic that says look
for analogies and stuff that I've learned before just like those Harvard kids the Harvard kids didn't even have a
heuristic that said gee let's see what do I know about the seasons absolutely
not they just started bullshitting that's exactly what they were doing because
that's what they were taught to do they never got any points anywhere for ever being able to come up with an analogy on
the spot with some prior knowledge that they had so when you're trying to do a
transfer experiment you immediately are to get transfer you have to build it
into the pedagogy to have the teachers ask the kids constantly for how is this
like that and how is this like that see here's looks let's end with this here's
a here's a really good way to think about this stuff most people can't
remember what it was like to learn as a child it's very hard to remember what it
was like but most everybody can learn what it was like to learn to draw a car if you recall this experience here
you are like this your mom is giving you advice which you
cannot hear there's this narrow tunnel down the road you can't remember what
gear you're in things are appearing in front of you you're over steering six
weeks later you're chugging down the road paying attention to things your
mom's talking to you're having a normal conversation maybe it's six months but
it doesn't take that long and here's what's happening cognitively is that
your brain your mentality is building little experts little simple experts to
take care of all of these tasks when you didn't know how to do it you were trying to do it with your sort of Center the
meanest the the conscious center that is
very inefficient at almost everything except talking and this is why it's hard
to learn to play tennis right off or hard to learn how to ride a car and
because the part of you that is this the meanest is basically one that knows how
to converse with other people but you can have a tennis pro tell you how to
hit a tennis ball until he's blue in the face and your body can't learn it from
the language statements some other thing has to happen and a lot of learning
especially this kind of learning requires the kids to build these little
modules that are essentially the heuristics that the scaffolders the
teachers and maybe sometimes even other kids can supply from the outside right
and I think anybody who's been to graduate school knows that a lot of graduate school is about internalizing
some of the heuristics from your gods the professors that you would like to be
more alike you're trying to find how do these people do this well this is what
it was like and when I was going to graduate school how do they do it and gradually you find here's a
professor that has a style it's kind of compatible with yours you start oh yeah
this works you get better at that and you wind up with a with a bunch of
little things that are constantly whispering to you like one of the great
heuristics is don't commit totally to
your first idea almost every kid does they're so happy
to have an idea they start following it and they can't ever remember that they
should really come up with a few ideas just in case so any adult that's helping
the kid can coach the kid to just do the thing okay let's you got here's an idea
let's get another one let's try a few things we'll just write them down and now the kid has a visual field of a
bunch of things I've done stuff with kids where I get the 5th grade class
this is what's interesting the 5th grade class you can get the kids to tell you
20 great heuristics for solving problems they know them they just can't use them
when they're solving the problems because they aren't part that's they
can't do it while they're in the midst of this stuff so a lot of what making
these things work is to have the heuristics in plain view somehow either
literally visually or through some scaffold ER who's just helping the kid
manage their attention to manage their
focus and stuff like that and after a while look just like driving the car the
kids start internalizing things that you can see it right in front of your eyes all of a sudden they get fluent like in
this in any computer system you've got usually three or four weeks of pain just
because you don't know what the conventions are there's too many things in the menus you don't know which ones
to go after and it's blowing your short-term memory out of the water just
to just to deal with the mechanics of it for weeks later the kids are just
blithely avoiding all of the minefields in the in the experimental software and
going to the places that work and doing this stuff that's great fun to do I mean
it's wonderful to do it and like I say if you can find the pace that both the
teachers and the kids are resonating with which is usually a slower pace than
any of us would want to do then amazing
things happen over even one semester the kids are in there's no way that you can
say that the kids aren't doing something like real mathematics because they just
flat are it's just not classical math
so unfortunately we're going to have to catch the buses the drivers are
threatening to leave okay