I believed until the age of 4
I remember very well
I was 4 and there it is on my bed and I
spoke with my grandmother and I told her
says if me put me in christmas green exists
we wait for me because the little
mouse I understand she needs
of time to make it his home have
so suddenly she comes back to recover our
milk teeth and as she was he
put the money under the cushion
so all that is very coherent and
him Santa Claus there is the kiss the
follies not because I did not see well
big but was in trouble I
continue to ask questions you
start acquiring his tour which
how fast
how many children in the world he has
I realized and I say them but
in fact Santa Claus is a party
news a few years later I am
student I’m 21 years old and I’m passionate
for a game the backgammon junks no
jonas and we play everyday together
and you know with the buddy at the beginning
we all play a game on them
it’s starting to be serious launches
charrie ans thing and then we play more
backgammon
we play the pad range
and when gammon day a moment
given I find myself in front of him have
said you can handle neither I saw an article
has passed a certain level
backgammon makes that lucky and
I was so disappointed because I
I had invested my friend Helen was
in the background he had created his own
party us because for me the kid
nice real strategy so I ran
at his home in a video library I
retrieved a book that was talking about
mathematics and german but I have them
it’s actually no it’s true that a
uncertain random part this is the
dice roll is a strategy and
then the free license at the end there
had an infection that suggested
many created an algorithm that plays
alone to the diamond I think I know
why it’s because I did not believe
not it will be better warning so I
am reading my book there is
a section at the end that talks about
machine languages
machine learning is the
possibility to create an intelligence
artificial that will achieve a
automaton an algorithm that plays at
balloon and so I code and I create this
artificial intelligence but there is a
particularity does that machine
learning
we do not explain all the rules
because it’s too complicated in the
machine learning is given that
examples I took examples of
part and I give them to
the algorithm after a moment
league at the rhythm started to
understand the rules so I told him
proposed to play with himself for
improve
then at the end the backgammon
the algorithm was really playing against me
and I was very proud because
I had been able
a moment to transfer no
from intelligence to intelligence
artificial for me it was the proof
dink backgammon
there is a part that is not a
some a part that is smart
since a robot was capable
mega man for me math
it’s like an anti foutage certificate
mouth when I have a problem is
that I do not know if it’s true too
this is wrong well then I try to solve
this question and suddenly I have a
better appreciation of the true and false
and I realized that I liked
solve these problems a few
years later I find myself at new
york and the encounter I come across a
sir who actually is a solid and
he is a very influential trader
very very powerful and I talk to him
I remember him he had a little the
Santa Claus look of the big ones he was
big and I told him jim what’s your
job there he explains to me he tells me
jérémy my job is to make
Applied artificial intelligence
to the stock markets to the king of the coup
I was hyper intrigued because I
I was interested in intelligence
artificial and I’m going to ask but
jim what is the point of using
artificial intelligence in the
stock markets so here he told me
an extraordinary sentence he told me
jérémy artificial intelligence
it’s the guarantee that it will not be
never any stock market crash
so I believed it and I worked with
finances and I made the intelligence
artificial applied to markets
fellows in 2008 in new york at
lehman brothers and suddenly a moment
given I felt that there was something
a little complicated so I left my
job and they sank is what in
reality they have never been able
to recover from my departure but to
me this story of telli jans
artificial that would be able to
guarantee the fact that would never again
no stock market crash it was it was a
party news but it was it was a
feast us too much it was the last
defeat us it was impossible for
me to support something like this
so I said to myself now
coming back to France to go to create a
company and their company has a
only goal is to be able
to explain to his clients what is
is true what is wrong with
discern the true
dufaux and never be
handle and well I can assure you
that there is work because in
company make the distinction between the
true and false it’s not easy
see all the colors
between those who are convinced that
artificial intelligence will kill
humanity and on the other hand those
who are convinced that intelligence
artificial doing miracles
it’s very complicated to tell them that
it’s possible it’s not possible
that you read it but you have not
including
but that’s true we can do it
right away and wrong things about
artificial intelligence
it can be read everywhere in Hong Kong a
law firm appointed in its edge
a robot fox news
researchers have been killing each other for
dozens of years tried to give
to artificial intelligence concept of
common sense and they claim they
that they gave the business sense
to make it impossible actually the news and
the truth is that they really did
but it’s a comic hit the robot
does not exist between them is not
able to have a real
discussion with lawyers from
understand the non- rational side
of a case the language hide that
can sometimes have when we talk about a
deal that’s complicated facebook has
created two artificial intelligences
these two artificial intelligences
open in a language that man does
do not understand facebook decided to
switch off
you’ve read this news fox news so
you are going to the bottom of the problem in
reality it’s true that facebook
works on natural language and has
precisely created two algorithms that
was supposed to converse in one language
natural to be able to negotiate but this
happened is that
experimentation failed because at
a given moment
instead of saying that they wanted
exchange of bananas they grasshopper
have started to say banana banana pom
pom pom and suddenly his volley bio Monday
he was not coherent he saw well
that the experiment was starting from scratch
then they did not leave but
for a moment we tried to
suggest that facebook
had gone to the brink of inventing
an artificial intelligence that
had plotted against the man and killed
humanity phew it was a time
French information works
with artificial intelligence the
msa bah no that’s true actually it’s
true and it’s a little scary because
we still think there may be
a subject of conflict of interest because
that if to work on the goals
against anti terrorism because real subject
we have to go through
algorithms that do not belong to us
not the command is able to justify
his with the French is theirs
give as we are able to protect
their personal data
then a subject of sovereignty of
algorithmic dependence
are we really want to
depend on someone for our struggle
antiterrorist are we not
able ourselves to create our
clean intelligent soft that’s it
a real subject the researchers have
recreated a brain in a computer
fox news because it’s impossible to
recreate an identical brain in
a computer
in fact what this article says is
something much more
particular the human brain it is
made of biological neural networks
neural networks that’s it
a system that allows you to transmit
the information of neurons in neurons
skis so that we are able
to discern seeing to feel of
to have emotions
the researchers they worked
on something else they worked on
a network of artificial neurons
it has nothing to see finally yes he
is inspired a little bit is a small
as if you were told that the
birds and planes it was the same
that would not come to mind to say
that researchers have recreated a
bird to make new york paris
no they took the idea a network of
artificial neurons c that’s a
system in each of the rounds that
represents a form of information and
each information is transmitted
between the Roma who are some to
side of the others at the beginning there are just
the pixels of a first photo a
photo of a car and as and when
measured by a layer system
artificial intelligence is going to be
able to understand from which car
we are talking
the branch today networks of
neurons that gives the most results
and that gives rise is the most of
research it’s called learning
Deep dip the ring into the dip
learning neural networks
artificial have the possibility to be
organized into layers tell themselves that the
very first he sees the whole picture
but the second of his current of
the image that through this
that art says the first and the
last they know only what is said
all the previous about
the image is where the intelligence
artificial will be created is that the
first layer must be able to
detect a notion of border of
perimeter contour then the second
category of shots I’m going to be
able to create a notion of form of
round of surface squares
and then the most noble layer the most
is able to recognize
real cars just to be
able to predict classifying and
tell which car is associated with
which marc this is it
truly what intelligence
artificial is able to do
today we give him a picture and
artificial intelligence will describe
what’s in the picture a woman
dressed in white with a black dress
in tennis kate on a green background with
two people who are behind
it is true what is wrong is hyper
important
it’s something very serious
because today we have the feeling
that we can hear a lot of things
on artificial intelligence but these
things we know we are not always
sure that they are true while in
reality and things that are very
beautiful and that are already realized
thanks to artificial intelligence and
who may be a little less
famous and those things that I’m going
tell you
I have a friend and doctors then you
know we say artificial intelligence
will replace the doctors
fox news by cons intelligence
artificial will help doctors this
enemy to the ball +15 because she is
both doctor and researcher in
biology she’s working on a
drug that is meant to give and
to return fertility to women who
have undergone chemotherapy
it’s a very serious subject and in
her studies she tests the drugs
on mice and to measure the impact
of the drug it measures in the
ovaries of smiles the number of cells
of follicles is what you know
how many of these girls are in
Mouse ovaries is what you know
how much there is in a mouse
finally you serve north is that
you know how much you have to study it
mouse to be able to test the
drug well do you figure that
charlotte she will spend 200 days at no
exclusively counting
mouse follicles in sections
of auvers for 200 days she will do
1 2 3 5
she will fill the number of follicles
that she saw in the first line I
I’m going to do that on the second one
200 days while she has a bac + 15 and
that she is a doctor would have come to see me
and told me I have a problem jérémy
nice I like the problems she
explained to me this counting problem
and I say come on we’re going to work
together on building a
artificial intelligence that is going to be
able to count cells at
early the first version was not
top
so do not find the same
results and then we were inspired by
work that and charlotte and then to
workforce today we have been
able to create an intelligence
artificial that counts the cells
she went from 200 days of work
ten minutes say it today
charlotte she plans to publish her
result of the communiqué of the
meet researchers to go to
conferences but even more she
plans to test another ten
drugs to go 10 times more
quickly in his medical research
this is a concrete example of what
artificial intelligence is capable
to bring
to the doctor that’s the real thing
often that artificial intelligence
is able to replace the trades is going
create one day a robot killer in
terminator
oh it’s not for right now
it is fact that news there is no
robots killers of swearing robots
of eurobonds firefighters a year ago I
met with lieutenant colonel of the
fire brigade of paris
he explained to me that he had a
owl problem
I like the problems he has me
explained that he has a system of
trucks but as he is in a
situation of budgetary restriction he
can not put 100% of the material in
each of the trucks so there are some
trucks that have all the equipment and he
are some trucks that are a bit of a
complete which in itself is not very
serious because when there is a
intervention he always sends the
truck that is complete but sometimes it
there are situations there are several
interventions and several emergencies
same time
it suddenly becomes dangerous because
that maybe he will not send the
good material on the good
interventions and that there are lives that
are at stake
so we worked together and at
more or less we have been able
to create an artificial intelligence
which according to the information he has
two types of calls the type
of intervention the type of distance the
key words that have come in he knows
to the advanced what kind of material it
is supposed to use and like that he
is not in a situation where he can
send an incomplete truck on a
intervention that is very urgent
it’s very concrete that it can
save lives jeans more warmly
colonel understood it something
he understood that math was a
certificate anti foutage of mouth of the
shot he went with our algorithm
he explained that he was not enough of
means he had more ways to
buy more equipment because he
had the mathematical proof
we also talk about intelligence
artificial that will create cyborgs
with news there is no trance
of death humanism from the death of
robots all this is not going to happen all
immediately the bionic man is not for
now
on the other hand, there are real achievements
concrete
we have been working for a little over a year
with
the senology team the hospitals of
strasbourg they came to see me in me
saying he had a problem
yeah I like the problems
they explained to me that after a
breast operation women who
so had breast cancer or a
very important risk of contracting a
unfaithful
I did not know what it was
the time a lymphoedema a little
of men this is a thickening of
arm but it’s an investment that
never returns to normal so in
actually it’s very unsightly and in
a framework of global therapy and shock
related to the disease to have in addition
bring something static innate
it’s complicated for these patients
we worked with them and we created
a sleeve connect a penguin that’s
like an armband that is threaded between the
arm and up to the shoulder
put it on the info but tree hill sleeve
we measure in real time
the circumference of the arm in several
places and like the chin it is
connected to his phone and that on
his phone one can know if he is
moving is a point
to play sports it’s a train of
do something else so the doctor he
is able to tell from the roof when you
play tennis with good regulation
of the lymph so suddenly that you
not cause lymphoedema inversely
when you do the computer and you
stay glued to your keyboard for
hours
the lymph circulates badly there is a risk of
Lymphedema that’s concrete
it’s true because today
it is a prototype ena in progress
industrialization
we will have to validate by laboratories
and tomorrow we actually will sell
the info m that will be worn by
people who will no longer have
problem from the side an aesthetics of
lymphedema that’s it’s true it’s
what artificial intelligence is
able to do so I have one last
example I do not know if I have time
for last example so I’m doing the
last example I have you know
today they eat the economic and
hyper important it’s called the
customer quality customer service it’s
the fact that we make sure that you are
happy you buy something
today when you buy on a
e-commerce site and that you type
something in the window
Well, there is someone who will help you
answer what time with your
parcel if it is lost or what they s
he
good size if there are any left
things in reality these people who you
answer they are in a place that
is called a customer contact center
they are all in front of computers there
will answer your pussy to respond to
your questions
the young customers the name is called
nathaniel and who came to see me saying
that he had a problem
your problem was a bit peculiar
is that the people who were in his
center of contacts they were they
were unhappy and in fact he is
sized very quickly because the
hyper job uninteresting because
the questions are always the same
the answers are always the same if
although in fact he did not really
does they have a response dictionary
ready made they know by heart the
Page and line number
they enter the number is entered and
they live like that and we make their
job is so uninteresting that
people are not staying long and everything
a galley for him to be able to
recruit people to do this
profession there so suddenly we have
worked with him we realized
that in reality there were 99 % of
conversations that was automatable
on which we could put a
artificial intelligence but this
artificial intelligence had the
ability to understand when the
conversation was too complicated
she had to call in a human and
blow the people who work
were much more interested
much more motivated and that makes it
to skyrocket the turnover
of this client there it’s a real
artificial intelligence achievement
it’s really concrete then you
tell me it’s really concrete
but there is also a stake a little
societal ethics because these people
there eventually they may go
lose their job and intelligence
artificial it’s true in part
participates in destroying jobs and this
which is really very important
that to be sure of whether we are doing
good choices to bad choices with
artificial intelligence is the
more important is to distinguish the
true of the false
because by saying that a white truth from
false
you’re sure you’ll never do
manipulated you are sure you do not
will never believe in defective in
thinking for yourself in
critical thinking
by learning by yourself you are
in a situation you will be able to
participate in deciding what is the choice
of life
that our society is going to have with
the advent of intelligence
artificial because that’s for sure
artificial intelligence will happen
and when she will arrive is better
know how to use rather
than to be forced to be manipulated
by people who would be bad
intentionally so I think back to this
discussion that I had with my
grandmother when I was four
when I told him I did not believe
to Santa Claus and today I’m there
think about it and I say to myself
artificial intelligence that delivers you
in 24 hours down from your home there
knows the history of all your purchases
who knows that if you liked such
product you will like such product it is
able to deliver all the France all
children in France in less than 24
evening hours of 24 November the 24
December
actually exists like what you see
it is extremely difficult to be sure
that we can distinguish the true from the false
thank you
[Applause]
”
Be First to Comment