more often lately you can hear
about obvious victories artificial
intelligence in 1997 chess
deep blue supercomputer won
harry’s strongest man chess
Kasparov in 2011 the car beat
man in his game and
2017 subdued poker its artificially
intelligence penetrates more and more
spheres of human activity now
cars are better than people on the exchanges
create synthesize new medicines
fly planes recommend movies
books music translate lyrics and so
further the American inventor and
futurologist ray kurzweil 5 said that in
2045 will be created artificial
intelligence that surpasses man in
all fields of activity that is
stronger in everything and since he is stronger
man he can create such
artificial intelligence which is stronger
that which can be created by man
there is stronger than himself and so on it
will continue until born
perfect such a computer god who
can all and has infinite
power is called this moment
called technological
singularity
many people are afraid of what supermind is
thinks that people are a hindrance that they
he is not needed and decides to get rid of us
how much is it likely and is it worth us
this fear lately all
more scholars agree that
waiting for technological singularity
worth deferring
and I will tell you why firstly
every artificial intelligence which
now exists
despite the fact that everyone calls them
equally skillful intelligence at the very
In fact, these are all different programs each
decisive decisive
narrow task in their subject area
eg
the computer that wins you in
chess cannot on this photo
to distinguish the queen from, say, Kasparov
the intellect that listens to your speech
and recognizes what you say yourself not
will be able to say a word and creation
in common
wide skillful intelligence which
combines all skills and all knowledge is
separate task
and scientists now do not even imagine how
to approach her
second, let’s compare the complexity of the games
in which the computer managed to beat
human and the complexity of our lives in
chess
in a typical move a chess player can make
one of about thirty options
action and all for the chess game
such moves
each of the two rivals does about
40
in the end, we get this number
combinations of possible combinations
chess games she is definitely a big
yes this is a unit with 120 zeros but still
it is much smaller than i am now
will show the example of human life rut
has much more variability as
you see the number of moves in the game more
number of different options to be like in each
the moment of time is also greater
and in the end we get a unit with 1000
zeros but let’s look at our
life at any given time you can
understand thousands of thousands of solutions now
do think what you can here in
this moment
for example do you can turn off or
turn on the phone you can hug
the person who sits on your right you
can escape from the hall funny waving
hands whatever thousands of actions and
even if you assume that such a decision you
take you say once a minute all
equal in life you have to
make 40 million such decisions and with
no rival plays you
and 7 billion people living on earth
further eighteenth year exposed problems
even in those areas that before
were considered
watch on artificial intelligence a
it was thought that everything was fine there
cars crashing robots
fall being unusual for yourself
situations google translates
meaningless texts in meaningful and
I’ll tell you such a thing until there are things
which are simple for man but
incredibly difficult for a car for example
let’s take the phrase
the boy took the cap off his head and threw it
everything is clear what happened you
painted this picture in your head a
the computer thinks like it is
maybe maybe something like this
actually this year in 2018
artificial intelligence organization
launched an artificial competition
intelligence algorithm are fed to the input
textbooks in various disciplines
exam questions drawn up on
basis of these textbooks every questions
4 answers given in this way
even a monkey or guinea pig class
I will lay and randomly answer
for every fourth question that is
stuff 25 percent what do you think
how much is gaining the best at the moment
artificial intelligence all now
people can create
geniuses creativity transcend monkey on
2 percent gains 27 percent ie
only two questions out of a hundred artificial
intelligence can two questions from then
can answer better than random
mechanism why is it that
some things
artificial intelligence are given under
easy and what is it to feel like
extremely helpless the whole point is
that intelligence is not one dimensional
you can not just jump and IQ
all suffered torn albert einstein
university graduate village
fool monkey dog pig
homer simpson there are many different
aspects of intelligence someone better
acts as a team someone alone
someone better get an exact science in
someone humanitarian
someone can concentrate more easily
stressful conditions
and someone calm is all aspects
intelligence and these aspects
also different in person and car
car easier
exact sciences and clear logic because
he has infinite computational
unlimited memory capabilities and here’s
it’s all man feels better in
flexible thinking we can simultaneously
combine knowledge with from hundreds of thousands
fuzzy facts that we call
common sense
we can all feel together
beautiful
beyond description logical
language we can receive
integrate signals from several
sense organs simultaneously and manage
hundreds of muscles and for us it does not deliver
no difficulties so maybe
it is worth trying the machine to repeat 1 in 1
human
and let’s use strong
the sides of each of them passed the first
tournament chess teams consisting of
human and artificial intelligence and
won in this tournament won not
team with the strongest computer or
not a team of the strongest chess players
won after mediocre cars
average chess players but this team was able
find the synergy factor then that plus
which connects the strongest side
man and computer here
looking for such factors more generally
sense we are engaged in our laboratory
hand paws and options are plenty for
start a person can train
artificial intelligence let’s
suppose we have a task
teach computer engino gram
determine the presence or absence of
human pneumonia for this engineer
should show computer set
frames with the disease without and explain where
the disease is present and where is absent
then at some point the computer will start
to think for yourself and bring help
to man
here everything would seem simple but even here
There is something to do because now
computer needs millions of shots to
learn to answer
and dozens of people we are trying to come to
the same number teach him for
isolated cases understand the logic when he
learned
the next is that everything can unfold
on the contrary, yes and now already artificial
intelligence helps a person imagine
imagine what a patient
who comes to the doctor
shows him in hologram and camera in
glasses doctor immediately determines
the presence or absence of the patient pneumonia
this patient has glasses
projected this state of health
patient his temperature pulse pressure
in to century in the office placed microphone
and who listen to the patient’s speech with the doctor
and automatically record form
history of this is called gain
intelligence and such systems can help
not only doctors but also a specialist in
most different areas
Can a car not only help
man answering the question but
coordinate a person’s actions or
maybe even a group of people for example
firemen in a burning building
imagine a system that
gets a picture from the fire brigade
connects together and build a map
burning room she knows where I was
present where it can offline
to calculate
spread fire and distribute
firefighters around the room in this way
to maximize the effect of extinguishing
fire to have minimize risk for
people if you combine all these said
situations then the picture turns out literally
such a hybrid enterprise
Imagine
The company is allowed to lead the clinic
which is a team man in
this clinic is assembled and coordinated
artificial intelligence though composed
of specialist people let’s say you need
take delivery or
the team is going to start the operation
the intelligence of each member of this team
supplemented amplified as I spoke
higher artificial intelligence and he’s in
turn
is trained by man such systems
this is a much closer future than
technological singularity is already
almost possible now but there is also
Problems
the fact is that the computer is too small
understands as I said before
it’s hard not to coordinate anything
people and we fight it recently we
conducted an experiment where they asked a group
people
answer the questions of that test with
exam questions from textbooks
by which I spoke before and
counted the number of correct answers
correct people answered 68 percent
questions then we wrote simple
simple program that chooses the answer
the one that gave the maximum amount
person
that is, simply selects the maximum and such
the program scored 71 percent correct
replies
after that we improved the program and
forced her to count voices
different professionals with different weight
for example if a person is often up
this was answered incorrectly by his weight
voting drops if he responded
correctly its weight rises and such
system
already able to score 81 percent correct
of answers then we are the weight of an expert who we are
spoke divided by areas of knowledge
for example, a person may respond poorly
biology and there its weight will be low but
respond well to computer science and there it is
the weight will be high we have done yet
several small improvements and in the end
have made the number 89 record percent
now
thanks now we are learning in this system not
just answer questions about explain
computer logic why the answer is why
the man came to this or that answer
build a logical chain we
we expect that in this case the computer
can sometimes give the correct answer
even when all people made a mistake
combining their knowledge and integrating them into
here we see what we see
a chain of knowledge from which to withdraw
correct answer we expect that
when we succeed we can yes
and what else is the most important intelligence not
just answer this question
specific he will also memorize these
knowledge gradually accumulating this
base here’s common sense
about which I spoke and sooner or later
learn how to respond without the help of people and we
we can increase that result 27
percent
up to tens of percent can
be 70 or 80 like here in this way
as I told you probably should n’t be in
Soon wait for that skillful
select crowds out people rather have to wait
that people will be reinforced by them
but it is not worth thinking that nothing
will not change of course there will be a shake how
and other technological revolutions
part of the professions will leave part will appear
new ones and those that remain can
change beyond recognition
accordingly what in my opinion will be
that what will be left to look at the routine
by computer
sincerely on he will take it upon himself
complex things such as science
art there people will be strengthened
and also the person will remain emotional
Aspects
This is a cinema and a theater of another Russian national team.
on football and there to help children in
old people and so on and at the end will tell
I have been programming history since early childhood
and when I was in school I thought that was not
I will live to the age when I go
to work as a programmer will no longer be
need all programs to be written
I understand now yes to me yesterday
turned 34 now I understand how I am
was wrong because things that can be
programmed just got bigger and
more it only seems that
artificial intelligence on the nose at the very
We are at the very beginning
big way
the next and depends on you what kind of
get this story you have all all
every chance to make her lovely thanks
[applause]
about obvious victories artificial
intelligence in 1997 chess
deep blue supercomputer won
harry’s strongest man chess
Kasparov in 2011 the car beat
man in his game and
2017 subdued poker its artificially
intelligence penetrates more and more
spheres of human activity now
cars are better than people on the exchanges
create synthesize new medicines
fly planes recommend movies
books music translate lyrics and so
further the American inventor and
futurologist ray kurzweil 5 said that in
2045 will be created artificial
intelligence that surpasses man in
all fields of activity that is
stronger in everything and since he is stronger
man he can create such
artificial intelligence which is stronger
that which can be created by man
there is stronger than himself and so on it
will continue until born
perfect such a computer god who
can all and has infinite
power is called this moment
called technological
singularity
many people are afraid of what supermind is
thinks that people are a hindrance that they
he is not needed and decides to get rid of us
how much is it likely and is it worth us
this fear lately all
more scholars agree that
waiting for technological singularity
worth deferring
and I will tell you why firstly
every artificial intelligence which
now exists
despite the fact that everyone calls them
equally skillful intelligence at the very
In fact, these are all different programs each
decisive decisive
narrow task in their subject area
eg
the computer that wins you in
chess cannot on this photo
to distinguish the queen from, say, Kasparov
the intellect that listens to your speech
and recognizes what you say yourself not
will be able to say a word and creation
in common
wide skillful intelligence which
combines all skills and all knowledge is
separate task
and scientists now do not even imagine how
to approach her
second, let’s compare the complexity of the games
in which the computer managed to beat
human and the complexity of our lives in
chess
in a typical move a chess player can make
one of about thirty options
action and all for the chess game
such moves
each of the two rivals does about
40
in the end, we get this number
combinations of possible combinations
chess games she is definitely a big
yes this is a unit with 120 zeros but still
it is much smaller than i am now
will show the example of human life rut
has much more variability as
you see the number of moves in the game more
number of different options to be like in each
the moment of time is also greater
and in the end we get a unit with 1000
zeros but let’s look at our
life at any given time you can
understand thousands of thousands of solutions now
do think what you can here in
this moment
for example do you can turn off or
turn on the phone you can hug
the person who sits on your right you
can escape from the hall funny waving
hands whatever thousands of actions and
even if you assume that such a decision you
take you say once a minute all
equal in life you have to
make 40 million such decisions and with
no rival plays you
and 7 billion people living on earth
further eighteenth year exposed problems
even in those areas that before
were considered
watch on artificial intelligence a
it was thought that everything was fine there
cars crashing robots
fall being unusual for yourself
situations google translates
meaningless texts in meaningful and
I’ll tell you such a thing until there are things
which are simple for man but
incredibly difficult for a car for example
let’s take the phrase
the boy took the cap off his head and threw it
everything is clear what happened you
painted this picture in your head a
the computer thinks like it is
maybe maybe something like this
actually this year in 2018
artificial intelligence organization
launched an artificial competition
intelligence algorithm are fed to the input
textbooks in various disciplines
exam questions drawn up on
basis of these textbooks every questions
4 answers given in this way
even a monkey or guinea pig class
I will lay and randomly answer
for every fourth question that is
stuff 25 percent what do you think
how much is gaining the best at the moment
artificial intelligence all now
people can create
geniuses creativity transcend monkey on
2 percent gains 27 percent ie
only two questions out of a hundred artificial
intelligence can two questions from then
can answer better than random
mechanism why is it that
some things
artificial intelligence are given under
easy and what is it to feel like
extremely helpless the whole point is
that intelligence is not one dimensional
you can not just jump and IQ
all suffered torn albert einstein
university graduate village
fool monkey dog pig
homer simpson there are many different
aspects of intelligence someone better
acts as a team someone alone
someone better get an exact science in
someone humanitarian
someone can concentrate more easily
stressful conditions
and someone calm is all aspects
intelligence and these aspects
also different in person and car
car easier
exact sciences and clear logic because
he has infinite computational
unlimited memory capabilities and here’s
it’s all man feels better in
flexible thinking we can simultaneously
combine knowledge with from hundreds of thousands
fuzzy facts that we call
common sense
we can all feel together
beautiful
beyond description logical
language we can receive
integrate signals from several
sense organs simultaneously and manage
hundreds of muscles and for us it does not deliver
no difficulties so maybe
it is worth trying the machine to repeat 1 in 1
human
and let’s use strong
the sides of each of them passed the first
tournament chess teams consisting of
human and artificial intelligence and
won in this tournament won not
team with the strongest computer or
not a team of the strongest chess players
won after mediocre cars
average chess players but this team was able
find the synergy factor then that plus
which connects the strongest side
man and computer here
looking for such factors more generally
sense we are engaged in our laboratory
hand paws and options are plenty for
start a person can train
artificial intelligence let’s
suppose we have a task
teach computer engino gram
determine the presence or absence of
human pneumonia for this engineer
should show computer set
frames with the disease without and explain where
the disease is present and where is absent
then at some point the computer will start
to think for yourself and bring help
to man
here everything would seem simple but even here
There is something to do because now
computer needs millions of shots to
learn to answer
and dozens of people we are trying to come to
the same number teach him for
isolated cases understand the logic when he
learned
the next is that everything can unfold
on the contrary, yes and now already artificial
intelligence helps a person imagine
imagine what a patient
who comes to the doctor
shows him in hologram and camera in
glasses doctor immediately determines
the presence or absence of the patient pneumonia
this patient has glasses
projected this state of health
patient his temperature pulse pressure
in to century in the office placed microphone
and who listen to the patient’s speech with the doctor
and automatically record form
history of this is called gain
intelligence and such systems can help
not only doctors but also a specialist in
most different areas
Can a car not only help
man answering the question but
coordinate a person’s actions or
maybe even a group of people for example
firemen in a burning building
imagine a system that
gets a picture from the fire brigade
connects together and build a map
burning room she knows where I was
present where it can offline
to calculate
spread fire and distribute
firefighters around the room in this way
to maximize the effect of extinguishing
fire to have minimize risk for
people if you combine all these said
situations then the picture turns out literally
such a hybrid enterprise
Imagine
The company is allowed to lead the clinic
which is a team man in
this clinic is assembled and coordinated
artificial intelligence though composed
of specialist people let’s say you need
take delivery or
the team is going to start the operation
the intelligence of each member of this team
supplemented amplified as I spoke
higher artificial intelligence and he’s in
turn
is trained by man such systems
this is a much closer future than
technological singularity is already
almost possible now but there is also
Problems
the fact is that the computer is too small
understands as I said before
it’s hard not to coordinate anything
people and we fight it recently we
conducted an experiment where they asked a group
people
answer the questions of that test with
exam questions from textbooks
by which I spoke before and
counted the number of correct answers
correct people answered 68 percent
questions then we wrote simple
simple program that chooses the answer
the one that gave the maximum amount
person
that is, simply selects the maximum and such
the program scored 71 percent correct
replies
after that we improved the program and
forced her to count voices
different professionals with different weight
for example if a person is often up
this was answered incorrectly by his weight
voting drops if he responded
correctly its weight rises and such
system
already able to score 81 percent correct
of answers then we are the weight of an expert who we are
spoke divided by areas of knowledge
for example, a person may respond poorly
biology and there its weight will be low but
respond well to computer science and there it is
the weight will be high we have done yet
several small improvements and in the end
have made the number 89 record percent
now
thanks now we are learning in this system not
just answer questions about explain
computer logic why the answer is why
the man came to this or that answer
build a logical chain we
we expect that in this case the computer
can sometimes give the correct answer
even when all people made a mistake
combining their knowledge and integrating them into
here we see what we see
a chain of knowledge from which to withdraw
correct answer we expect that
when we succeed we can yes
and what else is the most important intelligence not
just answer this question
specific he will also memorize these
knowledge gradually accumulating this
base here’s common sense
about which I spoke and sooner or later
learn how to respond without the help of people and we
we can increase that result 27
percent
up to tens of percent can
be 70 or 80 like here in this way
as I told you probably should n’t be in
Soon wait for that skillful
select crowds out people rather have to wait
that people will be reinforced by them
but it is not worth thinking that nothing
will not change of course there will be a shake how
and other technological revolutions
part of the professions will leave part will appear
new ones and those that remain can
change beyond recognition
accordingly what in my opinion will be
that what will be left to look at the routine
by computer
sincerely on he will take it upon himself
complex things such as science
art there people will be strengthened
and also the person will remain emotional
Aspects
This is a cinema and a theater of another Russian national team.
on football and there to help children in
old people and so on and at the end will tell
I have been programming history since early childhood
and when I was in school I thought that was not
I will live to the age when I go
to work as a programmer will no longer be
need all programs to be written
I understand now yes to me yesterday
turned 34 now I understand how I am
was wrong because things that can be
programmed just got bigger and
more it only seems that
artificial intelligence on the nose at the very
We are at the very beginning
big way
the next and depends on you what kind of
get this story you have all all
every chance to make her lovely thanks
[applause]
Please follow and like us: