Tip:
Highlight text to annotate it
X
>>Mary Lynn Broe: Welcome to the
fifth speaker of our 2011-'12
Visionaries in Motion Series,
Dr. Sebastian Seung.
It's good to see so many of you
here tonight supporting our
Gannet Project talks.
In the coming weeks, we will
have a web link featuring words
about the series for many of the
speakers over the past six
years.
You too can send us a comment
telling us what you think of the
Visionaries in Motion Series,
how it has inspired you.
Be sure to check our website for
both past years' videos and the
scroll of past speakers.
Now in recent years, the field
of neuroscience has grown in an
immense way.
From the popular Charlie Rose
series on the brain, the case
histories of Oliver Sacks and
the Nobel Prize winning research
on memory by Eric Kandel and the
language studies of Steven
Pinker.
Intellectual inquiry and
cultural curiosity about the
brain and its function is
widespread.
The Gannett Project has also
contributed a visible thread to
neuroscience at its intersection
with computational studies.
In 2006, the first year of our
Gannett Project, we opened the
cognitive revolution Darwin in
our everyday lives with Daniel
Dennett, philosopher and
cognitive scientist.
Dennett takes a computational
approach to understanding the
brain, arguing that human
consciousness and free will are
physical processes in the brain.
Other visionary speakers have
included Christopher deCharms,
founder of Omneuron who used
real time fMRI to look at the
brain activity to see how we
feel.
And of course in 2008 and '09,
AI expert and futurist Ray
Kurzweil was with us.
Tonight, Dr. Sebastian Seung
joins us to continue the
conversation with his
revolutionary work on the
connectome, the complete,
unique, wiring diagrams of
individual brain circuits.
That is to say the totality of
interactive connections among
neurons using new imaging
technologies in mapping neural
connections and synapses.
Professor of computational
neuroscience in the Department
of Brain and Cognitive Sciences
and in the Department of Physics
at MIT, Dr. Seung is also an
investigator at the Howard
Hughes Medical Institute.
He's also a member of the Max
Planck Society.
With a PhD from Harvard, Dr.
Seung worked at Bell
Laboratories.
And like so many of our
speakers, his research is fresh
and multidisciplinary standing
fields of physics, computer
science and neuroscience.
Dr. Seung has been our Sloan
Research fellow, a Packard
Fellow, a McKnight Scholar.
His research has been widely
cited in the New York Times
Technology Review, The Economist
and Scientific American and most
of you know his popular TED
video.
He's ongoing mapping of the
brain's wiring carries central
importance in understanding how
we learn, why we forget, as well
as an offering insights into an
autistic child, a dyslexic
sister or an Alzheimer's parent.
His new book, The Connectome:
How the Brain's Wiring Makes Us
Who We Are is available for
purchase and signing at the
reception after the talk.
As Dr. Seung discusses the
provocative new field of
connectomics, we might bear in
mind the comment of one
reviewer.
I think this was someone from
Scientific American.
The central question for Seung
and one that also keeps the
transhumanists on tenterhooks is
whether you are your connectome.
If you could deduce every
connection point of every brain
cell, the strength with which
every neuron fires and the way
these firing patterns change as
cells interact with each other
would in fact you be left with a
copy of you?
We hope two roving microphones
would be available for you to
ask questions and answers
afterward, if not, you're just
going to have to shout them.
Join us for a reception and a
book signing after the talk.
Pick up a postcard for coming
speakers.
The next one is a graphic
novelist, cartoonist,
illustrator, Charles Burns on
March 15th.
And tonight's caption talk,
together with today's lively
WXXI 1370 Connection chat with
Bob Smith will as usual be
posted on our website.
Please join me in a warm welcome
for Dr. Sebastian Seung speaking
on searching for the self in the
brain's connections.
[ Applause ]
>> Thank you so much, Mary Lynn.
That's a-- was a very eloquent
introduction.
Is the microphone working that
well?
No. Can we turn the volume up a
little bit?
>> Yes.
>> Thanks, Mary Lynn and thanks
Cassandra for organizing my
visit so nicely and I-- I've
really enjoyed meeting people so
far.
I-- we had a wonderful dinner
and I'm so glad that many of you
are enthusiastic about hearing
the latest that neuroscience has
to offer.
So how many people are we-- we
have in this-- we have a
distortion here.
[Background noise] Oops, sorry.
So I guess we do have some
technical problems already.
Let's see what we can do.
[ Pause ]
Alrigth. Alright.
So my book is launching and it's
got this strange sounding title.
How many of you have actually
heard of the term connectome
before?
It's quite a few.
You're on the-- your on the--
you're the *** for it, you know
the latest lingo that it's--
those of you who haven't the
word before, you shouldn't be
embarrassed because the word is
only I think 5 or 6 years old.
It was coined on only recently.
And there is a small group of
neuroscientists who are really
interested in finding
connectomes and I decided to
tell the story to the public.
You can't hear?
>> No.
>> Okay. How is this?
Any better?
Okay. I decided to tell the
story to the public and so I
wrote this book and I'm going to
try tell you something in my
personal excitement and why
we're doing what we're doing.
So Mary Lynn talked about why,
gave us-- gave you the story
that we're mapping the brain's
connections but why are we
mapping the brain's connections?
And often, it's more important
to know why people are doing
things than exactly how they're
doing them.
So I'll start out with that
first.
And I'll start with the question
that really neuroscience can't
answer yet but of course
everyone has thought about it
and has something to do with
neuroscience.
So does anyone recognize who
these two famous historical
figures are or were?
Does anyone know who these
people were?
It's a tough one.
What's that?
[Inaudible Remark] Ooh.
Anatole France.
>> Yeah. [Laughter] [Inaudible
Remark]
>> I thought-- I thought I was
going to be very impressed by
your-- about your man and
culture, but Anatole France,
okay.
So I'll give you the answer.
So on the right we-- on my
right, on your left we have Ivan
Turgenev, very famous Russian
novelist, author of Fathers and
Sons.
And on the other side we have
Anatole France.
Another fame-- really a beloved
figure of French letters.
And if you look at these men, I
think you'll notice that Ivan
Turgenev's head is very large
[laughter] it fills up most of
this oval right here.
Most of this he [inaudible] and
in fact, around the time that
this man lived, it was the
fashion to take the brains of
geniuses and remove them from
their skulls after they died and
for the learned professors of
the academies and universities
to figure out what was it that
distinguished the brain of men
of genius.
Of course they're always men in
those days.
It was a different world.
So, you can see that-- oh, more
technical problems.
Okay, I'm sorry.
So you can see that numerically
speaking, Turganev's brain was
twice the size of Anatole
France's brain.
In fact, Anatole France's brain
was substantially smaller than
the average brain which is about
1300, maybe 1400 grams.
And so when he died, there was a
famous British anthropologist
named Sir Arthur Keith who wrote
an essay about Anatole France
and he expressed his great
puzzlement.
He said in his great
professorial style, although we
know nothing of the finer
structural organization of
Anatole France's brain, we do
know that with it he was
performing feats of genius while
millions of his fellowmen
countrymen with brains 25
percent or even 50 percent
larger were manifesting the
average abilities of daily
laborers.
Quite staggering, isn't it?
And once, once I gave a-- I-- I
gave-- I read this quote at a
scientific meeting and this
French theoretical physicist
said, oh wait, he said, Anatole
France was not a great writer
after all.
[Laughter] But of course the
joke is that Anatole France won
the Nobel Prize in literature.
And in fact the-- Sir Arthur
Keith I think tried to-- tried
to sort of get out of his
puzzled state by making the same
conclusion as the French
theoretical physicist.
A detailed study of Anatole
France's life so far as is it
known shows us that he was in
many senses a primitive man.
So we laugh at this now because
we know that you really can't
predict the intelligence of a
person by the size of their head
or the size of their brain.
We know that that's really
silly, right?
And amazingly, this is the
normal range of human brain
sizes from 1 to 2 kilograms.
It's a huge range that you can
have a huge variation of brain
size and yet not much dependence
of intelligence on brain size.
And yet an interesting thing has
been found, which is that if you
look at statistical averages,
there's a weak correlation
between brain size and IQ.
So, if you look at people with a
very high IQ and you take the
averages of their brain sizes,
say 100 people and you do the
same thing with people with
lower IQ, you'll find that the
people with higher IQ on
averages have larger brains.
Now that study has been-- has
been done now with very fancy
brain scanner equipment.
You don't have to actually pull
the brain out of autopsy, you
can actually look at a living
brain and get very accurate
measurements of brain volume,
and has been shown that there's
this weak correlation, weak
dependence.
So in terms of statistical
averages, there is a dependence
of IQ or correlation between IQ
and brain size, but for any
particular individual it's
pretty much useless.
That correlation is so weak it
would be stupid to for example
give grades to my-- my classroom
based on people's brain sizes,
right?
That would be really dumb.
Okay, now what should we
conclude from this?
And I want to-- I want to stress
how primitive a state
neuroscience is in by showing
you this graph.
And so, we can make fun of
Anatole France and I get-- I
[inaudible] that as a story
about the 19th century.
But I want to assure you that
this is 21st century science.
So, here is an interesting
finding about autistic children.
It comes from a paper, you can
see down here the year 2005 and
the graph, the horizontal axis
is years of age and the vertical
axis is the difference between
the brain size of an autistic
child, average autistic child
minus the brain size of a
typical child.
And so what you can see is that
autistic children have a
substantially larger brain on
average.
So, you can see about 10 percent
larger than the typical child.
And this was in fact noted by
the-- the man, Kanner, who
originally defined autistic
syndrome in a landmark paper 60
years ago.
He noted that many of his
patients that he studied had
abnormally large heads.
So this is interesting because
first of all, it explodes the
idea that necessarily bigger is
better because it's true that
autistic kids may have special
talents but also they have some
disabilities.
And so the idea that brain size
is a good way of assessing
function gets turned on its head
in the case of autism.
Autistic kids have larger than
normal brains rather than
smaller than normal brains.
And it's more complex than that
because actually if you look at
them at birth, their brains are
slightly smaller on average.
And so there is a growth pattern
which is taking place and that
in adulthood, the size of the
brain is indistinguishable from
that of a normal population.
So again, we have we-- but we
have weak correlations.
If you just took every baby with
a large head and you said, oh,
that's an autistic baby, that
would never work.
Right. That's like saying that
any, any big person you see out
there is an NFL player.
It's true that NFL players are
larger than the typical person
but you can't, you can't turn it
the other way around.
These are statistical findings.
So this is I think showing you
at what an embarrassing state we
are in neuroscience that we
still have to rely on measuring
people's heads or brain size in
order to find out what's
different about an autistic
brain.
That's very crude and we have to
go, we have to go beyond that.
And in fact, this is something
that if you, if you look at the
history of mental disorder, so
let's say schizophrenia, there
has been 100 years of failure to
find any clear and consistent
pathology of the brain in
schizophrenia.
In other disorders, let's say
Alzheimer's disease, we know
that neurons are dying, right?
We know that some kind, some
kind of degenerative process is
taking place.
But in the brain of a
schizophrenic person, no clear
and consistent pathology, it
looks normal.
If you take the brain out after
the-- let's say this person dies
in a car accident, you take the
brain out, you look at it under
a microscope even, there's no
standard abnormality that's
associated with schizophrenia.
And not a lot of people are
aware of that.
They believe that science has
proven that schizophrenia is a
brain disorder, right?
But we had indirect evidence.
We have evidence that drugs can
help and so that suggests it has
a biological basis.
We have genetic evidence that
certain genes are correlated
with schizophrenia and yet we
have not found what's different
about the brains.
So neuroscience is far from
declaring victory in any of
these important issues.
We don't understand why people
are different.
We don't know what's different
about the mind of a Beethoven or
an Einstein.
We don't know what's different
about the brain of somebody who
has autism, schizophrenia,
bipolar disorder, and so on.
And so we resort to these things
like measuring brain size.
Now I'm criticizing brain size
but there's a positive spin that
we can put on it which is that
this suggest maybe we should be
studying the structure of the
brain, right?
Size is a crude measure of
structure.
So, some people might say we
shouldn't study structure at all
but maybe there are more
sophisticated measures of
structure that might tell us
more.
Now you may be aware of course
that in neuroscience today,
people do more than study
overall brain size.
They'll divide the brain up into
regions and here's a turn of the
19th century, early 19th century
map of the brain divided into
regions.
And we know that different
regions are activated when
different kinds of mental
functions take place.
And we, people have tried to
explain say, autism as the
result not of a larger brain
overall but an enlarged frontal
lobe.
Things like that, just taking
out one part of the brain and
focusing on it.
But what I want to claim to you
is that studying brain regions
is not enough.
We have to go further, we have
to actually study what brain
regions are made of.
After all, any of these regions
might consist of a hundred
million neurons.
And if we want to say why a
brain region works well or might
malfunction, we can't just leave
it at this level, we have to
take a reductionist approach and
we have to understand the basic
building blocks, the neurons of
which brains are composed.
And so in preaching like this,
I'm actually going back to my
roots as a physicist.
I studied theoretical physics
back when I was in graduate
school.
And one of the things I was
fascinated by of course was the
differences between different
kinds of matter and here is, on
the right, on this side on my
right you see a diamond and it
sparkles.
It's, as you know, very hard,
it's not conductive, it's an
insulator.
And over on this other side, we
see a piece of graphite, right?
Graphite is dark, it's soft,
it's a very soft material, and
it actually conducts
electricity.
And so a long time ago, our
predecessors would not have
known why it was and we can say
that this is different from
that.
But what's the fundamental
reason why these two are
different from each other?
And because of the achievements
of physicists we know that it
has to do with the atomic
structure.
So if you look at a piece of
diamond, people have found out
that the atoms are arranged in
this particular way so that
every carbon atom here is bonded
to 4 other carbon atoms around
it.
Whereas if you look at the
graphite, graphite is actually
composed of sheets of carbon
atoms, and each carbon atom is
attached to 3 other atoms.
And these sheets can slide past
each other and that's why
graphite is soft.
These sheets have different
electrical properties that's why
graphite can conduct electricity
and so and so forth.
So physics and physics, the
power of the reductionist
approach of breaking things up
into their constituents and then
understanding how these
constituents are organized, that
has been a very successful
approach and we need to do the
same thing for the brain.
Now one further point, of course
we can't take reductionism too
far.
We can't say that, oh, if I just
sit around on my lab and study
the properties of a single
neuron, then I will end up
understanding how the brain
works.
But I can't just study the
building blocks I've known, I've
got to study the whole assembly.
And how, how do we know that?
Well, look, the basic building
block in the case of diamond and
graphite is exactly the same,
it's a carbon atom.
It's not that the atoms are
different, right?
Carbon atom here, carbon atom
there.
If I study carbon atoms alone, I
would never understand the
difference between diamond and
graphite.
It's the arrangement, it's the
organization of the whole thing
that makes these two have very
different properties.
You know, if a physicist, you
know, a physicist are famous for
all kinds of things but if a
physicist were to-- a male
physicist were to propose to a
woman with a graphite ring and
say, well honey, it's made of
carbon atoms.
It's fundamentally the same
thing, right?
That he-- he would probably get
thrown out unless she were also
a physicist.
[Laughter] So that simple
example shows that it matters
how the building blocks are
arranged.
And so that's the difficult
part, so how can we apply the
strategy to the brain?
And I will assure you that it's
much more difficult to do this
for the brain than for diamonds
and for graphite.
And you-- and you might ask,
well you know, you've been
talking about this but why
haven't you guys succeeded yet?
And so, I'm making an excuse,
right?
I'm saying it's very difficult
for brains.
And today I want to give you a
little bit of sense for how
intricate it is, how the
arrangement of neurons in the
brain, their organization is
much more intricate than either
of these two mathematically
simple kinds of arrangements.
Alright, and I want to push the
idea of the connectome.
The connectome is the map of the
connections between the brain's
neurons, is one way of
characterizing that
organization, how the neurons
are connected into a network.
And I want to show you an
example connectome also.
The connectome that I'll show
you comes from this worm, it's a
1-millimeter long worm called--
it's called C. elegans which
biologists love to study.
It's got just 300 neurons and
its connectome looks like this.
Okay? This-- every node, so this
reminds me, you know, when
you're on the airplane and
you're really bored and you're
looking through the in-flight
magazine and-- [laughter] there
is the menu of all the air
food-- airplane food you don't
want to eat and some movies.
But then there's also the map of
the routes, right?
And, there're the cities that
the airline flies to and they
have 3-letter airport codes and
here we have a 4-letter neuron
code.
It's a-- turns out every neuron
in that worm has a name, a bunch
of letters like this.
And every line in this diagram
is not a route between two
cities but it's a connection
between two neurons, alright?
So this is a way of depicting
the network in a graphical way
and that this net-- this line
here indicates that these two
neurons can communicate with
each other because there is a
tiny junction called the
synapse.
The synapse is a junction in
which two neurons communicate
with each other.
So this-- this diagram has 7000
connections between 300 neurons.
And there is another diagram
that we could in principle draw
for each of your brains.
But it would be much more
complicated than this, right?
Because instead of 300 neurons,
you have about 100 billion,
alright?
So, 100 billion cities in that
flight map and here, 7000
connections with 300 neurons,
that's about 20 connections for
a neuron, right?
But you have about 10,000
connections for neuron.
So that it's really huge.
That diagram, there's no way
we'll be able to fit that on to
the screen with this number of
pixels.
It's a huge, huge set of
information, and that's your
connectome.
You've got one.
I've got one.
We all have one and there's some
similarity between them but
they're all unique.
They're all different.
And in fact, we're far more
different from each other than--
so worms, in worms, you can find
the same neuron in every worm.
There're 300 neurons and they're
arranged in the same location.
It's like a beautiful Swiss
watch.
Stereotyped, right?
And so that's why you can give
the neurons names.
But we can't do the same kind of
correspond-- one to one
correspondence between my
neurons and your neurons.
And as we saw before, Anatole
France and Ivan Tergenev, they
probably had very different
numbers of neurons.
So there's no way to put them in
one to one correspondence with
each other.
So there's a lot of variability
and I'd like to think that we're
more unique.
We have a certain uniqueness to
ourselves that worms do not.
This uniqueness of the
connectome is very important
because it reflects not only
your genetic inheritance but
also the experiences that
you've-- have influenced you
throughout your life.
One of the big hypotheses about
connectome is that your memories
are stored inside, right?
That's-- this diagram has a lot
of information in it.
Somehow, your memories have been
encoded in some pattern of
connections between your
neurons.
And one of the big challenges of
neuroscience is to figure out
exactly how that's the case.
And obviously, we've been
[inaudible] so far because we
can't map the connectome of a
person yet and this connectome,
although it's just 300 neurons,
this took over dozen years to
map.
It's a hard, hard work in the
'70s and '80s.
So, the big-- one of the-- the
new field of connectomics is all
about developing superior
technologies to enable us to do
this kind of mapping in animals
that had brains more like our
own and eventually maybe our
brains too.
Okay, so I have this slogan
which I-- in which I say you are
more than your genes, you are
your connectome.
Alright. So let me unpack that
statement.
I said, I emphasized the
importance of taking a brain
region and-- and following a
reductionist approach in which
we understand that brain region
has an assembly of a large
number of elementary units like
neurons, right?
Well, the reductionist approach
to the brain really works at
three different levels.
So neurons have activity, right?
You've heard that electrical
signals can travel around the
brain.
Every neuron is actually
producing electrical signals
whenever you're thinking or
doing things.
Well, not every neuron.
So it's actually every neuron
can, some neurons talk at some
times, other neurons talks at
other times.
It actually is a complicated
kind of symphony.
So, neural activity is a thing
that people, scientists see when
they do brain scanners.
They can see whole areas of the
brain light up.
You've seen those kinds of
pictures in newspapers, but we
can also do more sophisticated
refined measurements that
measure the activity of a single
neuron, not the overall
activation of 100 millions
neurons.
So neural activity, there's a
lot of evidence that your
current thoughts and your
feelings, your perceptions,
they're all encoded in this
complicated signal in between
neurons.
And then I emphasized the notion
of connections, right?
So, the connection between
neurons are really important
because they determine which
neuron talks to which other
neuron when it-- when it sends
these signals out.
And more importantly than that,
they also determine in a way
that I've-- it's more intricate
but I've explained it in my
book.
They determine how neuron
computes.
How it takes its inputs and
decides-- and decides when to
produce a signal.
And then finally, inside--
buried inside the nucleus of a
neuron is-- is the DNA, the
genes.
And so these are the three
fundamental quandaries that we
have to come to terms with if we
want to understand the brain.
The activity of neurons, the
connections of neurons, and the
genes, they all have a-- they
all have their role to play at
any understanding we're going to
have about how the brain works.
Now why do I emphasize one of
them?
Why is my book called Connectome
and not Brain Genome or-- why am
I emphasizing the Connectome?
And the reason is that-- so I--
I truly believe that we have to
understand the brain in all of
these levels but I believe that
the Connectome is the
fundamental one that we really
care about because it
corresponds to the notion of
personal change.
So genes, of course, don't
really have much notion of
change in them, right?
Your-- your genome was created
at the moment of conception, and
it essentially has not changed
since that time.
So the genetic idea, certainly
we know that genes influence our
traits.
They have a lot of-- they have a
lot of-- of power over the way
that we are.
But genes don't change.
On the other hand, if we look at
neural activity, that's
ephemeral, that flickers like
this, right?
So it's going on and off, on and
off.
Let's say you get mad one
instant then you get reflective
and you realize you shouldn't
have gotten mad.
Your-- your mental states are
changing all the time and neural
activity is changing all the
time like that.
And the connectome actually
occupies a middle ground.
The connectome is the wires of
the brain, the branches, the
branches of neurons.
You know neurons are shaped like
trees and they have these long
branches.
Those are pathways along which
electrical signals can flow.
Those wires of the brain and the
connections, the synapses
between those branches, they are
much more stable than electrical
signals, right?
Wires are something that you can
think of as a stable thing.
They're real-- real material
solid structures and they stick
around.
And so the connectome I would
argue corresponds to what we
think of when we think of the
self as a stable quantity.
Right, day after day, you-- you
come in and-- or you wake up and
you see your spouse or your
kids, and they're roughly the
same person they were yesterday.
It may-- I guess sometimes you
might be unhappy about that.
But that-- [laughter] Right?
So-- so your-- [laughter] So,
they're roughly the same person
and that's because there's a--
they have-- there's a stable
notion of self and I would argue
that's the connectome.
The connectome can change.
It can change slowly over time
but it doesn't change rapidly
the way neural activity does.
So the connectome really
determines what neural activity
patterns get expressed.
What is-- how a person will
react to a certain stimulus, and
so what is that for?
So the connectome is more
dynamic than the genome.
So it allows for the prop-- the
possibility of personal change
but is it more stable than
neural activity.
So it accounts for-- for why a
person was-- why we are fairly
stable and often why we can't
change even if we want to.
And so that brings me then to--
the question of course, the
people always ask which is the--
so ultimately when you talked to
people about neuroscience, I
think fundamentally what they
care about is the possibility of
change, right?
They want to know, can my
grandmother recover from her
stroke?
Can I quit smoking?
Can I learn how to be a more
compassionate person?
All these things are very hard
to achieve and they are
probably-- if this is correct,
they're most probably about
changing your connectome.
So, that's why I believe that
the connectome is fundamental--
has a fundamental place in
neuroscience research.
It's about knowing your self and
about finding new means of
change in your self.
That's what we want to find out
from neuroscience.
And scientists have already
found the mechanisms by with
connectomes change.
So I call them the 4 R's,
reweigthing, reconnection,
rewiring and regeneration.
You could think about the brain
as a vast jungle of these 100
billion trees that are all
extending branches and
reweigthing means a change in
the strength-- the size of a
synaptic connection between two
of these branches.
Reconnection would mean true
creation and elem-- or
elimination of connections, and
rewiring would be actual growth,
extension of branches or
retraction.
And regeneration would mean
creation and elimination of an
entirely new neurons, the
creation of entirely new neurons
or destruction of existing
neurons.
And we know that all of these
processes take place some more
than others as-- some less than
others as you get older.
And so these 4 R's, we want to
understand how exactly they're
related to all these kinds of
personal change that I
discussed.
And in the end, we would like to
know how to promote them, to
affect the changes that we want.
And that I submit would be the
mission of neuroscience and
that's-- it's these mechanisms
that change the connectome.
This is-- I'll just site two
very-- sort of well-known things
in popular culture, I think that
talk-- they kind of give an idea
how much we are ambivalent about
the notion of change.
So here's Rob Reiner around
1998, there was the-- this idea
of the first 3 years, be really
important and Rob Reiner made a
video call "The First Years Last
Forever" and this was made to
push the idea that it's really
important to ensure that our
children have really good
stimulating environment,
nurturing environments, because
the effects of the early years,
of the first 3 years, last
forever throughout life.
And so you should really
concentrate your resources on
the first 3 years.
And I'm not going to argue
against that, I certainly
believe that it's important to
give children stimulating
experiences.
But on the other hand, look at
this, on the right hand side,
2007, is another book called
"The Brain That Changes Itself"
by Norman Doidge, "Stories of
Personal Triumph from the
Frontiers of Brain Science".
And this book describes all
these case studies of adults,
people who suffered brain
injuries.
The doctors wrote them off as
unable to recover.
And this, these people were able
to make this inspirational
recoveries and Doidge argued for
the sort of infinite capacity of
the brain to remake itself, to
rewire itself, to use the 4 R's
to heal.
And you can see that these two--
these are actually not
compatible ideas, right?
This, this one is saying that
adults can't change and
therefore, we have to make every
effort with the kids, and this
book is saying that adults can
change in whatever way they want
it, they can do these miraculous
things.
But this is the problem that has
to be attacked through science
and neuroscience is really not
there yet.
So, we know that the truth can't
be that totally and we know that
it can't be this totally, it's
somewhere in between and we need
a more sophisticated
understanding of that, that can
only come from deep and rigorous
scientific study.
[ Pause ]
Alright, so I've set up the
problem and I want to say that I
will now talk about the approach
that we're taking which is to
actually look at the brain at
the level of-- at a-- using very
powerful microscopes and to look
at the branches of neurons and
to actually trace the
connections, to see how neurons
are connected in reality of the
brain.
And once we can actually see
those connections, map the
connections, we would have a
chance to really understand the
4 R's.
And one interesting thing that's
happened is that we have come to
a crossroads at which in order
to proceed, I need to ask all of
you to get involved.
That there's going to be a way
for the public, not just
specialized neuroscientists to
do this kind of research, but
the public will be able to
participate in mapping the
connections of the brain.
And so that's what I'm going to
talk about now in greater
detail.
So how do we do that?
Well, here is actually a famous
neuroscientist that--
neuroscientists love to add
their heroes.
This is one of the heroes,
Santiago Raman y Cajal.
He won the Nobel Prize in 1906
for his discoveries of neurons.
You can see him there with his
microscope and he made many
beautiful drawings.
He was like a neuron collector.
He would peer into his
microscope at these pieces of
brain tissue and he would draw
these pictures like this.
This is sort of early
illustration, he was-- so his
contributions was he was-- he
was-- he can be called the
discoverer of the neuron.
It was very hard to see neurons
until his competitor, Camillo
Golgi, an Italian physician,
invented a special method of
staining neurons so that only a
small percent of the neurons
were stained.
So before, imagine how this
dense jungle of neurons, he
can't see anything because the
branches are all so tangled up
together, it's just a big mess.
But now, suppose that 99 percent
of those trees become invisible,
only 1 percent of them are
stained with this dye, that's
what Golgi achieved and Cajal
used that to discover what
neurons are truly shaped like.
They have this cell body and out
of them come all these intricate
branches.
But that's not enough to study
connections because here we're
only seeing a few neurons at a
time.
If we want to find a connectome,
we need to see all of the
neurons together, we want to see
the-- every tree.
And for that, we can use very
high resolution microscopy.
And so I'm going to show you a
little piece not of the brain
but it's a piece of the retina,
so that the neural tissue at the
back of the eye-- so you may not
know this, but at the back of
the eye is a little piece of the
nervous system with neurons that
actually process visual
information.
And you're going to see a block
of tissue that is as wide as a
hair as thick, so extremely
small.
And it's a 3 dimensional block
of tissues so you'll see the
slices peeled away from that
block to give you a view of the
3 dimensions and that's the gray
scale images.
And then you'll see a branch of
a neuron being traced through by
a computer.
We have the sound?
[ Music ]
So what you just saw, slice by
slice were cross sections
through entangled spaghetti of
neural branches.
All those brown circles you saw
in every slice, they were the
cross sections of neurons
created by cutting through a
very, a very, a very thin slice,
you know, a thousand times
thinner than a hair, using the
world's sharpest knife, a
diamond knife.
So the color branches that you
saw were traced by computer.
A computer, artificial
intelligence developed in my
laboratory was able to
essentially do what a kid could
do, color in the areas inside
the boundaries that you saw and
trace out the path of the neural
branches.
And that's credible task because
if you want to figure out how
neurons are connected to each
other, you have to trace the
pathways of the wires to find
out what's connected to what.
Now the problem is that
computers are still not perfect.
So as you all know you've seen
robots in science fiction movies
that can see but in reality
robots don't see nearly as well
as humans do.
So our AI still makes mistakes.
And so humans have to oversee
the operation of the computer to
guide the AI.
And it's very, it's very, it's
still very time consuming to do
that and so what we like to do
is enlist the power of the
people, have people come and
help us interact with the AI and
help us map the connections of
the brain.
And so we're launching a website
and I'll show it to you here.
[ Silence ]
>> It's call Eyewire, I like a
E-Y-E like the eye, and the
tagline the call at the bottom,
it says welcome the Eyewire
where you can help make
discoveries about the neural
structure of the retina.
And this tab right here explain
why we want to study neurons,
what we can learn from studying
the retina, and they get some
instructions and I'll take-- and
this actually, those are
actually neurons of the retina,
and I will take you now to the
play tab.
[ Silence ]
And what you can see right here
is a little piece of that image
you just saw.
So not the whole thing but even
smaller, a little manageable
piece like this.
And at the top, right here, is
where the computer has been
coloring, and if we now
navigate, we can go slice by
slice through where the computer
has been colory.
And we can see, I think you can
see the computer has stopped
prematurely.
Can you guys see that?
So right here, the computer
should have colored further and
so we can now help the computer
out by clicking right here and
the computer then knows, it's
not totally dumb, the computer
is able to then color, keep on
coloring and complete most of
the rest of this object.
And on the left hand side here
you can see a 3D view.
So the computer is actually, you
and the computer are interacting
to actually coloring this small
piece of a branch of neuron, and
this yellow plane is the image
that's been shown right here,
okay?
We can inspect that to see
whether the computer has left
anything else out.
You can see that it's still a
bit sloppy.
If we're perfectionist, we might
go here, and make a few more
clicks, throw it in.
And this is a tutorial so in
order to teach people this task,
we set it up so that.
In this case, an expert is
already done this task, and
you're compared against an
expert, if you click right here,
that's wrong, right?
So you can get some feedback,
immediate feedback.
And there's a progress bar and
we're trying to do things to
gamify this, to make it fun.
It's already, somewhat
hypnotics.
So some people just like
watching this.
It's kind of hypnotic, right?
You know some people just don't
like this at all and other
people like my girlfriend,
she's-- I'm kind of worried
about her, she's addicted.
[Laughter] I think-- she yeah,
some people question our
relationship, but she really
enjoys doing this.
She thinks it's very relaxing at
night.
It kind of calms her down and
it's like, you know, kind of
like a crossword puzzle.
It kind of occupies your mind
but it's not too, it's not
disturbing anyway just something
smooth like this.
So we're trying to make this as
fun as possible so people will
treat it like a game.
They'll get rewards, points and
so on and we want to make it a
fun activity for everyone to
come together and help us map
the connections of the brain and
also to learn about the brain at
the same time.
So this is the part of a new
trend called citizen science to
get everyone involved, everyone
working together and do
something exciting.
Oh yeah, I want to tell you that
this is just the first project
of an organization that we're
starting called "Wired
Differently", uniting us to
explore the jungle of the mind.
And the retina is just the first
stop.
We want to actually study
memories so we want to test the
hypothesis that when you
remember something, when you
store a new memory, your neurons
get wired differently by that
experience.
And then we want you test that
connect all of the hypothesis
that conditions like
schizophrenia or autism, are
actually somehow associated with
abnormal patterns of connections
between neurons.
And we're going to enlist the
public to help us search for
those abnormal patterns of
connections.
And just to drive that point
home [pause] you know, look at
this picture.
Look at the nightmare, right?
So can anyone guess, what's the
total length of wiring in your
own brain?
Any guess?
>> Several thousand miles.
>> Several thousand miles.
Several thousand miles is
actually more, I think the total
length of blood vessels on your
brain is probably-- [Inaudible
Remark] So the estimates would
be millions of miles of neural
branches inside your head.
Think of the possibilities from
missed wiring, right?
I mean if you just look at
behind your stereo [laughter] If
I were really malicious, I would
just go switch 2 wires like that
you wouldn't never know the
difference just looking at
casually take and look at the
wires.
You would have to trace the
connections, right?
You got to trace the wire, where
did they go?
And that's the kind of problem
we're faced with in finding
connectomes.
And this is just shows you what
happens if you analyze the wires
in many such small cubes and
assemble them all together, you
can get an entire neuron which
looks like this.
That's from the retina, this is
from the cerebral cortex.
And the dendrites-- many of the
neurons in the cortex are
studded with these torn-like
protrusions called spines on
which excitatory synapses get
received.
So to conclude I'd like to say
that I'd like to conclude with
sort of a philosophical question
that people love to ask, right?
Is the brain complex enough to
understand itself?
It's like, oh yeah.
[Laughter] You got me.
So that's one of those
philosophical questions.
It's really tough but what I
would submit to you perhaps
millions of brains assisted by
the most advanced artificial
intelligence we have can do the
job, and so I'd like to ask you
all to join and spread the word
to help out.
So let's take some questions.
Yup!
>> Is it visible to imagine
stimulating a lot of these
connections and see what
happens?
>> Right so neuroscience had
done that for a long time.
In fact I was a theoretical
physicist and I became a
theoretical neuroscientist and I
stimulated the brain.
I tried to make stimulations
parts of the brain.
But then I realized, well, these
stimulations are so constraint.
We have to know how the neurons
are connected to each other to
have a chance of stimulating
them.
So you may have heard of the
blue brain, Henry Markram's
project in Switzerland.
He claimed he got to stimulate
an entire brain but he doesn't
have a connectome.
So you have to-- I think to have
a realistic brain stimulation
you have to know the connectome
first.
Now will the connectome be
enough?
That's a big question.
So I discussed that in the final
chapter of my book and I take
the-- I discussed in fact that
question not in the scientific
manner but in a manner which
takes on a whole kind of I guess
it call it the rupture of the
nerve.
So the fantasy are that people
to live forever.
Can I upload my brain to a
computer and live forever?
Right, it's the bid for
immortality and so that's, you
know, I got some heavy going in
the book in the middle, but in
the end we have a little fun
with the more sort of fantastic
science fiction aspects of what
could be done with the
connectome, ah yes
>> So [inaudible] cognitive
functions and emotion being all
connections and activities or is
there other aspects that are yet
to be done or not.
>> Well there are-- so our
emotion is fundamentally
different from thoughts for
example, like Mr. Spock versus
Bones, okay.
[Laughter] Alright, well maybe
in some ways different brain
areas are involved and maybe
different neurotransmitter
systems but, you know,
neuroscientist in the end would
say that it is about spikes, you
know electrical signals and
chemical secretions.
Now we haven't proven that but
that's-- there's a lot of
evidence for that, lot of
evidence.
There are definitely areas of
the brain that are very
important for emotional
recognition and expression.
Do we have a question at the
back?
[ Inaudible Remark ]
>> Great, great.
Alright so the question is about
glial cells, and so glial cells
are nonneuronal cells in the
brain.
We don't talk about them as
much.
They're-- so neurons are like
the stars of the show and glia
are like the crew, right?
The show can't go on without the
crew.
But they're not the stars of the
show.
So there's a lot of evidence
that glia are usually important
for brain diseases.
If your glia go bad you're brain
is not going to function right.
But I've taken the strong
position here that, that sort of
what we think of it as thoughts
and perceptions, that's
primarily the sequela neurons
and there's a lot evidence for
that idea.
But, yeah, glia are certainly
important.
Okay, we have a question there.
>> Here you're talking about the
brain function and what I'm
wondering is you have input and
if you're thinking of change or,
you know, what if were thinking
of change, you get input, a
variety of inputs, perceptual or
inner movement.
>> Yes.
>> And these are going to affect
your function and then you have
an output that you haven't even
need.
My question is if you clear but
I think you catch the idea of
input and output from your
lecture.
>> Well.
>> [Inaudible] a little bit.
>> Well, the question of input
and output is important but what
I'm primarily concerned with, is
how does, how the past input,
how did that past input leave a
trace on the connectome?
Alright, so the problem of
perception may primarily be sort
of, how do we-- what's happening
in the brain due the current
input.
But the problem of memory is
really about what happens if
that left a trace in the past
and that's what connectome is
really very important for.
Now you ask about loops, the
fact that when I send something
then I do something and what I
do influences what I sense next.
So that's important, that's a
big loop.
But in fact that's a relatively
simple loop compared to the
loops that are inside your
connectome.
So there're all kinds of
incredibly looping pathways that
are circulating inside the
brain.
[Inaudible Remark]
>> That's right.
That's right.
So, in fact your key-in on one
of the most difficult problems
in neuroscience which is, how to
understand all these circular
path wave?
If you take biology class they
love to draw these linear steps
of metabolic pathways like, but
really in biology there's always
some kind of thing that goes
back.
And it's a difficult problem and
technologies are now getting to
the point where we can start to
take a part these loops and
figure out what's going on.
>> Thank you.
>> We got a question here.
[Inaudible Remark]
>> Well, I don't know, but what
I try to communicate today is
that anything you do could
conceivably be changing your
brain structure.
So it's-- the fact that-- so in
the theories of neuroscience
whenever you remember anything,
all the memories that you store
from yesterday, the reason you
are able to remember them is
that some kind of trace got left
in the connectome.
So it's not just, it's not
meditation.
>> We got a question here.
>> Yeah, you've been talking
about the connection within the
brain itself, but do you have
any theories about brains
between brains.
Like not necessary psychology
but do you think there's any
technology or business behind
it.
>> Connections between 2 brains.
Well, I mean the obvious
connection between 2 brains is
language, right?
But I guess you're asking
whether there's something about
that is beyond the realm of
physics.
Yeah I don't know of any
evidence for it.
[ Laughter ]
I don't have any idea.
[Laughter] But maybe I'm just
attuned, I don't know, I'm not
attuned.
I'm not attuned, but yeah good
question so yeah.
[ Inaudible Remark ]
>> Well, certainly there are I
have colleagues that are working
on spinal cord.
I even have colleagues who are
working on the connections
between the spinal cord and the
muscles.
We are-- the retina is kind of
on the periphery, but actually
in the end, I'm more interested
in what's happening in the
brain.
The brain is-- so the difference
of course between us-- one of
the differences is between us
and worms is that our neurons,
most of our neurons really are
concentrated inside a particular
place inside the head whereas in
a worm the neurons are much more
distributed across the body.
So, so yeah, but they're all
interesting questions, they're
all interesting questions.
Yeah, over there in the--
[ Inaudible Remark ]
>> Okay, so you're bringing up
the question of epigenetics and
that's actually something I
didn't touch on at all, and it
is a huge, an interesting
research topic in neuroscience.
So epigenetics of the brain,
just sorry, I don't-- it's
really fascinating but I just
didn't have time to discuss it.
[Inaudible Remark]
>> Oh, yes, oh yes.
So, I said genes are essential
to any account of the brain and
genes can influence your brain
in many different ways, and one
important way is by controlling
the process by which your
neurons wire up, so genes shape
the connectome, definitely.
Yeah?
>> [Inaudible] of the worm.
Have you ever tried or anyone
tried to stimulate actually how
the brain in the worm
[inaudible] And did the
simulation matches to what you
see in reality [inaudible].
>> Some people have tried, but
what's missing from the worm is
that we don't know how the
individual neurons work.
So strangely enough, it's more
difficult to record electrical
signals from a worm neuron than
from a mouse neuron or even a
human neuron.
Just for, we have technical
reasons, it's hard to do, and so
we have almost no idea of the
electrical signals that are
going inside that network.
So we have lots of information
about connections but no
information about, almost no
information about signals.
You got question?
[ Inaudible Remark ]
>> Well that the-- okay, so the
analogy, so one of the popular
analogies of the brain is to a
transportation or communication
network.
And the only function of such a
network is to get, to be able to
route a plane or a signal from
any point A to any point B
inside that network.
And so that why that's a
misleading analogy for the
brain, I mean, there are some
notion to it, right?
So there's some notion that if I
step on a nail that information
has to get to my vocal chords so
I go ouch.
So it is to a certain extent,
it's true that my nervous system
is communicating.
But the difficult thing about
the nervous system is
understanding why a particular
signal here ends up producing a
particular output.
So if I see a snake, I might
turn to run away, if I see a
steak, my glands start to
salivate.
So why doesn't the steak make me
turn away, right, turn to run
away?
Why does-- in both cases the
signals are coming into my eyes,
the input, and yet they have
very different effects.
And so that's not just the
communication problem, there's
really a computation that has to
take place.
And the connectome tells more
about, tells a lot of-- in my
book it explains why people
think the connectome says a lot
about the computations that
neurons perform, not just
communication pathways.
Question?
>> Right, so suppose we observed
missed wiring that's associated
with schizophrenia.
Supposed that we were able to do
that, we don't really know if
that's the cause still, and so
ultimately the only way to test
that would be to alter the
wiring and see what the effects
are.
So that's another hard problem,
right?
And it's the same thing with
infectious diseases, right.
So, I might observe a microbe
that is always associated with
the symptoms of disease, but I
still haven't proven that's the
cause until I kill the microbe
and see what happens.
So ultimate you have to do that.
But a strong correlation, a very
strong correlation is often a
good indicator that there may be
a causal relationship.
That's why size is useless
'cause it's such a weakly
correlated thing.
Way in the back?
[ Inaudible Remark ]
Yes, so it's a difficult.
It's going to be a difficult
computational problem.
Once we have connectomes.
So it's a same way where
genomes.
So one of the things that people
do is compare genomes with
different individuals or
different species and that
actually relatively simple
computation because it's just
one dimensional sequence.
But a connectome is a graph, a
map like this and so, computer
scientists are going to work on
algorithms for doing those
comparisons, it's gonna be a
whole new field.
We got a question there.
>> So do you think on this study
the computation with connectome
would allow us to understand
consciousness or subjective
feelings 'cause it seemed to me
I'm more than just my
[inaudible].
>> Right! So this is the problem
of a subjective experience of
consciousness, quality and so
on.
No, I don't think the connectome
is really about that.
That more has to do with the--
so there's, there's, there's
several motion of self, and so I
talk about the stable self
that's the connectome.
But then there's a self that--
that changes all the time,
that's the one that associative
with the consciousness and with
normal activity.
So it's really a different--
different kind of question.
I think it's an interesting
question, but I don't know how
to answer it, yeah.
[ Inaudible Remark ]
Well it's a snap shot, it's a
snap shot, right?
So, you're-- so the notion here
is that your self is changing
slowly every time and you're
taking a snap shot that's
capturing yourself at one point
in time.
And so what we're trying to do
is understand the encoding of
your memories or whatever part--
you're properties at this
particular point and time in
your connectome at that
particular point in time.
[ Inaudible Remark ]
That we can't, okay, that-- that
we can't do because this method
of finding connectomes rest on
dead bricks so this is one
measurement.
But we could study different
conditions, right?
So what affect does if we have a
mouse and they put it in rich
environment and another mouse in
deprived environment we could
see what the differences is in
their connectomes.
>> If you took two genetically
identical mice and supposed
that's going to take additional
[inaudible] whether that's, you
know, [inaudible] of food and
water or you know, more of--
paralyze them 'til it can't
move, you know.
You know that has the same, you
know, some input.
Do they happen [inaudible].
No. At a certain level yes, but
why, because I-- we should just
back up even with worms their
connectome is not exactly the
same.
There's slight, the neurons are
all the same but the set of
connections between neurons is
even slightly different from
worm to worm, and there's even
more differences for people.
[Inaudible Remark] Well, they
just a lot of-- okay, it could
be random noise.
There're a lot of small random
events that happen during
development.
You could think about, imagine
we have two contractors who are
both building a house according
to the same plan by an
architect.
Will those two houses end up
being exactly the same?
There'll be differences,
there'll be differences.
Now imagine something that's
incredibly more complicated than
house.
Alright?
[ Inaudible Remark ]
>> Would this is basic research
but I would hope so because what
I'm trying to emphasized here is
the ability to see what's wrong.
If we can see something
different about the brain in
pathological conditions then we
can try to find drugs that
eliminate that difference,
eliminate that abnormality and
we can even see that right now.
So imagine that you want to look
for drugs that are good for
infectious diseases but you
couldn't see the bacteria.
It's a lot harder because you
can't see the effect on cause of
the causative agent.
So seeing better is a first
step, it's not a solution by
itself but I think going to
help.
Do you have a question?
[ Inaudible Remark ]
>> Yes, so this is a--
[laughter] People's from here,
know.
So this is a, this is a whole
research subject now, right, the
idea of the Wisdom of Crowds.
So, if you have the same task
done by 5 people and you use
that redundancy to detect the
errors, you might end up with
something better than an expert,
right?
But indeed the, the whole
question of how do you use
computers to put together
people's mind power and achieve
a better result than any single
person alone, that's an
important frontier in computer
science.
Over there?
[ Inaudible Remark ]
>> So we will, meet apart from
that but that's really is a
central activity that we've put
this, we take this, we take the
image, the entire image, we chop
it up into little blocks and
then we farm out those blocks to
many people who do it as a
gaming kind of activity.
And through that, we could
imagine doing the image analysis
much faster-- so just for
reference.
So if I took the images of a
cubic millimeter brain at that
resolution, and supposed we have
no assistance from AI, which is
to have one person, coloring
painstakingly every slice after
slice.
So don't laugh because there's a
post doc in my lab who has done
that [laughter] but not for a
cubic millimeter for much
something much tinier.
And so, if we extrapolate form
that, he's actually a German so
his, he doesn't [inaudible] He
doesn't any [inaudible].
And if we extrapolate from that,
a cubic millimeter with take
between 100,000 to million years
to color in, if you never slept
or take coffee breaks, okay.
So we need AI to speed it up but
AI still wasn't good enough.
Let's say AI makes it a hundred
times faster, a thousand times
faster.
Then we need lots of people to
work, get a little bit of their
time.
And not only that so those
people, so we use an approach in
AI called machine learning.
So as the people put their
decisions into the computer, the
computer learning from those
decisions, they're becoming
smarter.
So we need all of these things
in order to succeed in analyzing
this huge amount of data.
Yeah, right there.
[Inaudible Remark] [Laughter]
>> As soon as possible but this
in participation.
How many people will
participate?
So, have you heard Galaxy Zoo?
Galaxy Zoo is a Citizen Science
Project to classify galaxy--
images of galaxies from Hubble
Space Telescope, you could look
it up online.
I think a hundred and fifty
thousand people signed up in the
first year and they classify
something like 15 million
galaxies or something like that,
okay.
Those are pretty easy judgments,
you get one picture and you just
press the button, say whether
there's a spiral or something
likes that.
This is more complicated visual
task but, you know, I think
brains are more exciting than
galaxies personally.
So I want to get a million
people to sign up and help out.
Yeah?
>> If you could be thinking that
connectomes are essentially
mapped with behavior patterns or
most probably [inaudible].
Does it all [inaudible]
behavioral patterns and
[inaudible] to seek finance.
How is your work at all useful
and applied to today's living
say [inaudible] in understanding
that?
What if I will tell you, would
it be better to actually
[inaudible] function studies
from a behavioral or the most
biological basis or any basis to
actually understand how the
brain works?
So for example what you can
[inaudible], you'll see that
It's like a [inaudible] and
that's actually a lot more
useful in understanding how to
treat that disorder, how to
understand it, how to understand
your own thought to go into that
and through that, understand the
thoughts of--
>> What-- that's not
historically how it happened.
Historically what happened was
the drugs through serendipity,
we discovered they had good
effects on psychosis.
And then it was discovered that
those drugs had some effects on
the dopamine system, right?
So I wouldn't call it-- I
wouldn't really call it-- it's
not even clear that it's
fundamentally a dopamine
disorder.
That's just one theory, there's
also the glutamate hypothesis,
and so on and so forth right?
So it's not-- that kind of
understanding from drugs is
really I would say very
primitive.
I'm not-- it's the best that we
have but there's no reason to be
satisfied by that.
Alright, so really those-- that
understanding comes from
serendipitous discovery that
this chemical compounds have an
effect and it's a theory that's
made up after the fact to
explain why the drugs have their
effect.
Okay?
>> Well, because reduction is
the purchase of work really well
before.
I consider infectious disease,
right?
That you should go below the
surface level of symptoms and
you find something in the unseen
world that has an effect that
causes the symptoms, that's a
very powerful discovery.
We haven't gone to that point in
neuroscience yet, right?
And so some people may be
skeptical that we'll ever get
there but I would still hold out
hope that we will get there.
Right, so drugs-- antipsychotic
drugs are purely discovered on
the basis of symptoms and I'm
not going to deny it.
Those are the most useful thing
we have right now but we want to
go deeper than symptoms.
[ Inaudible Remark ]
>> There's going to be many
levels, right?
So for example look at the
genome right?
So mapping the genome has been
enormously important but it's
sort of just the beginning.
It points us into the right
direction where we can ask the
right questions right?
So just because you can see
what's wrong with the brain, it
doesn't mean you could fix it.
There're plenty of genetic
disorders, you can get a test
for certain disorders like
Huntington's disease, right?
You can get a genetic test that
will predict what basically 100
percent certainty whether you'll
get that disease later on.
So we know the cause, we can
test it but we have no cure.
But it's got to be a step in the
right direction, right 'cause we
found the causative agent.
Okay, so I'm not claiming that
everything will be solved by
finding a connectome but I'm
claiming it's a step in the
right direction.
Question? Yeah.
>> If you go to the case of the
words, again, is it possible to
follow the connections as they
occur when it's in a few cells
and then watch it progress and
maybe infer rule?
>> So people are starting to do
that now.
So the problem was that finding
that connectome took a dozen
years over dozen years and it
was so traumatic that no one
want to do it again.
But now that things have
gotten-- technology has gotten
better, there're some scientists
that Albert Einstein calls
medicine who are doing more
worms and they want to do
different ages, they're doing
different sexes, et cetera, et
cetera.
So that's going to happen.
Developmental-- how connectome
develops over time is a huge
usually important question.
.
[ Inaudible Remark ]
.
>> Well, I think the [inaudible]
is used for forming out
computation to people's
computers, right, and that's not
necessarily what we need here.
We actually need people's
intelligence, what's up here.
Not what's in their PC at home
because we have all kinds of
powerful computers which aren't
smart enough to solve the
problem all alone.
We need human intelligence and
machine intelligence together.
I think the hour is getting late
so why don't we--
>> Reception?
>> Yeah, why don't we, we could
go to the reception.
We could go home and go to
sleep.
[ Laughter ]
[ Applause ]