this is the kind of stuff that makes you
fall in love with neuroscience he is
beautiful stunning complex and intricate
and this is just one neuron out of 80
billion neurons and each and every one
of your brains the man who drew this
picture in the 19th century the father
of neuroscience Santiago Ramon Inka Hall
famously quoted that every man if he so
desires becomes the sculptor of his own
brain how could we possibly master this
complexity my name is Nikolai I’m a
neuroscientist and I got my PhD from
this place here and spent the last 10
years working on the question of
mastering our brains from various angle
looking at how to make neurons how the
neurons find each other how they connect
with one another and what happens when
those connections were those connections
in diseases like neuropathic pain and I
can tell you mastering our brains isn’t
an easy question and quite paradoxically
the more we dig into this question the
more difficult it appears and I’m not
exaggerating about its complexity if we
talk about the place where neurons
connect with each other where they
communicate this place is called the
synapse and here in red circle it’s just
one of those in that neuron and if we
zoom into it
this is what it looks like it’s a mesh
full of proteins each one performing its
function critical for our survival each
neuron has between a hundred to 10,000
of those synapses that makes trillions
of synapses in each human brain those
synapses they determine who you are how
you feel what you do and what you’re
about to do
and if this doesn’t look complex enough
imagine for a second that these sinuses
are constantly changing even as you’re
listening to this talk and learning
about sinuses now if that’s the first
time in your life you heard about
sinuses and it’s never gonna come up
again the connections that have just
changed in your brain they’re eventually
going to disappear and wither away but
if I reiterate that the place where
neurons interconnect is called a synapse
that there are trillions of synapses in
our brain that it is their scale and the
ability to change that largely
determines who we are I’ve just
strengthened those connections so to
that extent we’re able to sculpt our
brains but what happens when these
connections do not function properly
when we cannot control our attention
when we cannot control how our sleep in
insomnia when we cannot relax and then
in more severe cases when the sinuses
get hyper activated and this results in
epilepsy or they get under activated
results in mental retardation or when
the neurons can connect with each other
to support each other and then that
results in neuronal death in diseases
like dementia Alzheimer’s Parkinson’s
for example the fact that we still don’t
have an adequate treatment for those
diseases gives us a glimpse into how
little we know about our brain that we
cannot master it let alone engineer it
what we can engineer however our
artificial brains or artificial neural
networks there is no pretty picture and
the concept is very abstract so what I’m
going to do is I’m going to represent a
neuron with a palm of my hand and you
can imagine five numbers going down the
down the hand into the into the palm
where they get processed and the palm
decides whether to send a number down
the arm and here there’s another neuron
connected to this one rather awkwardly
and this strength of the connection then
determines which number gets sent down
here
and then if you put three of those
together in a particular way with
particular strength of the connections
they can perform simple logical
functions and then you start putting
more and more and more of those neurons
and they start performing more complex
functions but how is it possible to know
how to connect those neurons what
strengths should they be connected at
and this is where mathematics comes in
it’s just beautiful formulas comparing
the output of that network to your
desired output and if it’s far from
expectations then the formulas just
change those weights and change and
change until the network gives you the
desired output and hopefully it can then
go into real data and provide you
meaningful outcomes this is how learning
works in artificial neural networks and
it has some success in the early days
with image recognition recognition of
speech or writing but the real
breakthrough came very recently when it
was possible to scale up that network
and then use all the vast amount of data
available for training it and this
became known as deep learning and it had
its applications in image processing
that went beyond the capabilities of
what we can reach in defeating master
players at a game called go at doing
better algorithms for drug optimization
and many many many more applications the
list is growing the complexity of those
networks however is nowhere near the
complexity of our brain so at this point
you’re probably wondering how are the
two concepts are introduced so far
connected the artificial intelligence
and the biological intelligence and I
will tease you with that question for a
moment before but before I wanted to
introduce a personal story there was a
conversation with me and my longtime
friend Darin he is a diplomat and an
interpreter and a filmmaker and I was
around the time when I found out
but Elon Musk’s neuroscience venture and
said Darren your well-connected could
you put me in touch with Elon either
directly or indirectly and she said yes
why I said well I I want to work for him
okay but why well because I have you
know I admire this person for his great
vision and inspiration and I also have a
couple of ideas that I think could
benefit his company it’s starting to
look a little bit like a job interview
so what happened is next what happened
next is not something that ever would
happen to you at a job interview he said
Nikolai why don’t you take these ideas
and then build your own company could it
wow that was a bold and pivotal question
I mean yeah sure I have a chance and you
know in today’s world where you know
North Korea is shaking hands with
America the word impossible just loses
its meaning so we started a company
called it brain patch and the idea of
the company was to connect brains and AI
this was an insurmountable challenge for
the two of us so we started looking for
help and we have spoken with different
people first of all to our common friend
professor sheffield university professor
buzz back the flat oaf who gave us some
initial advice and provided us with some
initial guidance then got a financing
perspective from a Cambridge based
venture capitalist Max talked to one of
friends of one of old friends of mine
who is now back at Imperial doing
research in deep brain non-invasive
stimulation and I think his research is
so exciting it just deserves this whole
separate TED talk
dr. nian Grossman connected with
Professor Steve Porter who is an expert
in neural engineering and artificial
intelligence he’s already given his TED
talk
serendipitously connected to professor
paul Fitzgerald who is done 20 plus
years of non-invasive brain stimulation
and last but not least spoke with Sergei
Petrovich who was a blockchain he was a
blockchain
expert and was working on it long before
Bitcoin became cool and many many other
people so our venture was starting to
look a little bit like a neural network
with you know those those connections
and I think this is what any
organization should look like a network
of people connected by a common desire
to produce a certain output and capable
of modifying those connections in that
structure this is important if the
output is far from desired so let me
explain to you how something that seemed
a bold and borderline crazy idea turned
into something that was starting to look
like a feasible project and this
revolves around the idea of simultaneous
recording and stimulation of the brain
so recording you have heard in the
earlier talks is something that has been
available for half a century and
recently people use it for using used
brain waves to control wheelchairs
drones virtual objects etc and then
non-invasive brain stimulation it’s been
known since about a century ago when
Galvani discovered electricity people
tried zapping their brains with
electricity because why not and then
eventually now with the emergence of new
technologies where you’re able to
specifically targets particular areas of
the brain this is now using a new
revolution in brain stimulation so with
that ability to simultaneously stimulate
and record from the brain we’re able to
see the effect of that stimulation in
real time so it means that if the effect
of that stimulation is far from the
desired effect
we can apply learning to modify the
stimulation parameters and this is
exactly the kind of task that I’ve just
explained can be sold by artificial
intelligence in real time that
optimization of stimulation for each
individual for each application for your
brains and it might seem a little bit
eerie at this stage especially given all
the negative connotations that happen
with the artificial intelligence in the
media but I think maybe 10 20 years down
the line you’ll be sitting there and
thinking well what is brain without the
eye and the answer to that is simple
it’s just Bruner so at this at this
stage you’re probably thinking so now
what what about it so you’re able to
connect brains in the eye what’s the use
of it I’ll give you just a teaser a
simple example this is a recording of
brain activity in normal awake states
this is by the way my required recording
from from my brain a part of the
recording and basically using
alternating current stimulation we’re
able to change that and by research has
shown that by optimizing that current to
the individual frequency of each brain
you’re able to achieve a state that
looks a little bit like this and what
this is is the state of a brain in when
it’s relaxed so by doing this
closed-loop non-invasive stimulation
using artificial intelligence to
optimize this for your brains you’re
able to achieve a result that you
couldn’t achieve sometimes it’s
difficult to achieve without the
stimulation so if this was a product and
I believe that
the money should go from the consumer to
the developer only once that particular
state of relaxation has been reached and
this is a perfect case for the use of
smart contracts that have recently
emerged it also gives you an opportunity
to store this information of the success
of the application on an immutable
distributed ledger acting a little bit
like clinical trials so that delivers a
safe effective and trustworthy
application of non-invasive brain
stimulation and this is just one use
case currently the uses of non-invasive
brain stimulation combined with
recording are limited by our
understanding of what we can what
brainwaves mean and what we can record
from them and this can have implications
in diseases like in everyday life for
relaxation as I have mentioned there is
potential to do it for sleep for
concentration for learning but it also
can have an effect on clinical side
there is potential for treating epilepsy
there is potential for treating
Parkinson’s disease with technologies
that are going deeper inside the brain
depression is already being treated with
non-invasive brain stimulation and
there’s even scope for investigating its
effect on Alzheimer’s disease and I
believe that by having this
decentralized and democratized were able
to take advantage of all the vast amount
of knowledge from everybody on the
planet this way delivering this platform
and putting you in charge thereby
envisioning the prophecy that every man
if he so desires
can become the sculptor of his own brain
thank you you
fall in love with neuroscience he is
beautiful stunning complex and intricate
and this is just one neuron out of 80
billion neurons and each and every one
of your brains the man who drew this
picture in the 19th century the father
of neuroscience Santiago Ramon Inka Hall
famously quoted that every man if he so
desires becomes the sculptor of his own
brain how could we possibly master this
complexity my name is Nikolai I’m a
neuroscientist and I got my PhD from
this place here and spent the last 10
years working on the question of
mastering our brains from various angle
looking at how to make neurons how the
neurons find each other how they connect
with one another and what happens when
those connections were those connections
in diseases like neuropathic pain and I
can tell you mastering our brains isn’t
an easy question and quite paradoxically
the more we dig into this question the
more difficult it appears and I’m not
exaggerating about its complexity if we
talk about the place where neurons
connect with each other where they
communicate this place is called the
synapse and here in red circle it’s just
one of those in that neuron and if we
zoom into it
this is what it looks like it’s a mesh
full of proteins each one performing its
function critical for our survival each
neuron has between a hundred to 10,000
of those synapses that makes trillions
of synapses in each human brain those
synapses they determine who you are how
you feel what you do and what you’re
about to do
and if this doesn’t look complex enough
imagine for a second that these sinuses
are constantly changing even as you’re
listening to this talk and learning
about sinuses now if that’s the first
time in your life you heard about
sinuses and it’s never gonna come up
again the connections that have just
changed in your brain they’re eventually
going to disappear and wither away but
if I reiterate that the place where
neurons interconnect is called a synapse
that there are trillions of synapses in
our brain that it is their scale and the
ability to change that largely
determines who we are I’ve just
strengthened those connections so to
that extent we’re able to sculpt our
brains but what happens when these
connections do not function properly
when we cannot control our attention
when we cannot control how our sleep in
insomnia when we cannot relax and then
in more severe cases when the sinuses
get hyper activated and this results in
epilepsy or they get under activated
results in mental retardation or when
the neurons can connect with each other
to support each other and then that
results in neuronal death in diseases
like dementia Alzheimer’s Parkinson’s
for example the fact that we still don’t
have an adequate treatment for those
diseases gives us a glimpse into how
little we know about our brain that we
cannot master it let alone engineer it
what we can engineer however our
artificial brains or artificial neural
networks there is no pretty picture and
the concept is very abstract so what I’m
going to do is I’m going to represent a
neuron with a palm of my hand and you
can imagine five numbers going down the
down the hand into the into the palm
where they get processed and the palm
decides whether to send a number down
the arm and here there’s another neuron
connected to this one rather awkwardly
and this strength of the connection then
determines which number gets sent down
here
and then if you put three of those
together in a particular way with
particular strength of the connections
they can perform simple logical
functions and then you start putting
more and more and more of those neurons
and they start performing more complex
functions but how is it possible to know
how to connect those neurons what
strengths should they be connected at
and this is where mathematics comes in
it’s just beautiful formulas comparing
the output of that network to your
desired output and if it’s far from
expectations then the formulas just
change those weights and change and
change until the network gives you the
desired output and hopefully it can then
go into real data and provide you
meaningful outcomes this is how learning
works in artificial neural networks and
it has some success in the early days
with image recognition recognition of
speech or writing but the real
breakthrough came very recently when it
was possible to scale up that network
and then use all the vast amount of data
available for training it and this
became known as deep learning and it had
its applications in image processing
that went beyond the capabilities of
what we can reach in defeating master
players at a game called go at doing
better algorithms for drug optimization
and many many many more applications the
list is growing the complexity of those
networks however is nowhere near the
complexity of our brain so at this point
you’re probably wondering how are the
two concepts are introduced so far
connected the artificial intelligence
and the biological intelligence and I
will tease you with that question for a
moment before but before I wanted to
introduce a personal story there was a
conversation with me and my longtime
friend Darin he is a diplomat and an
interpreter and a filmmaker and I was
around the time when I found out
but Elon Musk’s neuroscience venture and
said Darren your well-connected could
you put me in touch with Elon either
directly or indirectly and she said yes
why I said well I I want to work for him
okay but why well because I have you
know I admire this person for his great
vision and inspiration and I also have a
couple of ideas that I think could
benefit his company it’s starting to
look a little bit like a job interview
so what happened is next what happened
next is not something that ever would
happen to you at a job interview he said
Nikolai why don’t you take these ideas
and then build your own company could it
wow that was a bold and pivotal question
I mean yeah sure I have a chance and you
know in today’s world where you know
North Korea is shaking hands with
America the word impossible just loses
its meaning so we started a company
called it brain patch and the idea of
the company was to connect brains and AI
this was an insurmountable challenge for
the two of us so we started looking for
help and we have spoken with different
people first of all to our common friend
professor sheffield university professor
buzz back the flat oaf who gave us some
initial advice and provided us with some
initial guidance then got a financing
perspective from a Cambridge based
venture capitalist Max talked to one of
friends of one of old friends of mine
who is now back at Imperial doing
research in deep brain non-invasive
stimulation and I think his research is
so exciting it just deserves this whole
separate TED talk
dr. nian Grossman connected with
Professor Steve Porter who is an expert
in neural engineering and artificial
intelligence he’s already given his TED
talk
serendipitously connected to professor
paul Fitzgerald who is done 20 plus
years of non-invasive brain stimulation
and last but not least spoke with Sergei
Petrovich who was a blockchain he was a
blockchain
expert and was working on it long before
Bitcoin became cool and many many other
people so our venture was starting to
look a little bit like a neural network
with you know those those connections
and I think this is what any
organization should look like a network
of people connected by a common desire
to produce a certain output and capable
of modifying those connections in that
structure this is important if the
output is far from desired so let me
explain to you how something that seemed
a bold and borderline crazy idea turned
into something that was starting to look
like a feasible project and this
revolves around the idea of simultaneous
recording and stimulation of the brain
so recording you have heard in the
earlier talks is something that has been
available for half a century and
recently people use it for using used
brain waves to control wheelchairs
drones virtual objects etc and then
non-invasive brain stimulation it’s been
known since about a century ago when
Galvani discovered electricity people
tried zapping their brains with
electricity because why not and then
eventually now with the emergence of new
technologies where you’re able to
specifically targets particular areas of
the brain this is now using a new
revolution in brain stimulation so with
that ability to simultaneously stimulate
and record from the brain we’re able to
see the effect of that stimulation in
real time so it means that if the effect
of that stimulation is far from the
desired effect
we can apply learning to modify the
stimulation parameters and this is
exactly the kind of task that I’ve just
explained can be sold by artificial
intelligence in real time that
optimization of stimulation for each
individual for each application for your
brains and it might seem a little bit
eerie at this stage especially given all
the negative connotations that happen
with the artificial intelligence in the
media but I think maybe 10 20 years down
the line you’ll be sitting there and
thinking well what is brain without the
eye and the answer to that is simple
it’s just Bruner so at this at this
stage you’re probably thinking so now
what what about it so you’re able to
connect brains in the eye what’s the use
of it I’ll give you just a teaser a
simple example this is a recording of
brain activity in normal awake states
this is by the way my required recording
from from my brain a part of the
recording and basically using
alternating current stimulation we’re
able to change that and by research has
shown that by optimizing that current to
the individual frequency of each brain
you’re able to achieve a state that
looks a little bit like this and what
this is is the state of a brain in when
it’s relaxed so by doing this
closed-loop non-invasive stimulation
using artificial intelligence to
optimize this for your brains you’re
able to achieve a result that you
couldn’t achieve sometimes it’s
difficult to achieve without the
stimulation so if this was a product and
I believe that
the money should go from the consumer to
the developer only once that particular
state of relaxation has been reached and
this is a perfect case for the use of
smart contracts that have recently
emerged it also gives you an opportunity
to store this information of the success
of the application on an immutable
distributed ledger acting a little bit
like clinical trials so that delivers a
safe effective and trustworthy
application of non-invasive brain
stimulation and this is just one use
case currently the uses of non-invasive
brain stimulation combined with
recording are limited by our
understanding of what we can what
brainwaves mean and what we can record
from them and this can have implications
in diseases like in everyday life for
relaxation as I have mentioned there is
potential to do it for sleep for
concentration for learning but it also
can have an effect on clinical side
there is potential for treating epilepsy
there is potential for treating
Parkinson’s disease with technologies
that are going deeper inside the brain
depression is already being treated with
non-invasive brain stimulation and
there’s even scope for investigating its
effect on Alzheimer’s disease and I
believe that by having this
decentralized and democratized were able
to take advantage of all the vast amount
of knowledge from everybody on the
planet this way delivering this platform
and putting you in charge thereby
envisioning the prophecy that every man
if he so desires
can become the sculptor of his own brain
thank you you
Please follow and like us: