a devourer of novels books
science fiction books
of science and among my authors
favorite
there was stephen jay gould and douglas
hofstadter had no equal
for intermingling in their history the
adventures of humans the adventures of
ideas the adventures of big projects
art culture culture science everything
together
Stephen Jay Gould made us dream
with evolutionary biology stories
my great love of douglas youth
hofstadter lead us into
incredible mathematical adventures it’s
in these works I heard
speak for the first time of alan
turing the development of
artificial intelligence and all
other kinds of topics there is the
mathematical and the rest and I think that
it played a role in the fact that I
become a mathematician later but
the chance of studying has made me
am found mathematician not at all
on the topics of water predilection
flattering but on other things
exciting mathematical physics
in front but what’s his darling
mathematical physics problem of the
stability in plasmas problem of
the growth of the company in the
very strong subjects gas well away from
the image of blur that had intelligence
artificial at this time
and when one of my colleagues went
launched in there in intelligence
artificial in the 2000s I say
what you sure about what you do is a
subject that it did not result in much
for now and as I’m wrong and
everyone wrong person
did not anticipate that it would come back on
the front of the scene with so much strength
these are the hazards and rebounds of
science and then I too had my
dose of going to two rebounds
the unexpected story and a few years
later I found myself deputies in
charge of a report for the government
on busy artificial intelligence
to delve into the subject and listened
hundreds of experts telling us about
evolution of this topic intelligence
artificial it’s exciting because
finally it’s much more effective
that we believed him but especially his
is invited in all fields is a
transversal topic this a topic in
which one must listen to the experts from
countless horizon a subject of which he
must be discussed publicly so that the
society can appropriate the evolution
to invite situations of
rejection to allow everyone
to be a player in this evolution
when we think about intelligence
artificial immediately images
scary come to mind
we think whether machines and
algorithms and will come to rob us of our
jobs it’s an intelligence
superhuman will seek to take our
control like in the stanley movie
kubrick 2001 the space odyssey
we can fear if as the
mathematician quétier o’neal that these
fearsome weapons can be used at
goal service that we can only
to disapprove and we have many
examples in which a bad
use of artificial intelligence
can succeed
to a destruction of democracy and this
are fears that need to be addressed
posed while trying to do everything
what it takes for them to get
do not realize artificial intelligence
it is also a subject that invites itself to
very positive way in our dreams in
what we are able to achieve
in incredible achievements like
the one that led two miss abyss to
start building the best
go player never realized we’re there
let’s get the alpha go program through
statistical techniques that now
make the heart of the methods
of artificial intelligence thanks to
reinforcement techniques that
allow the algorithm to
perfect in the directions that
have never been tried was developed
an algorithm that you are stronger than
any human go player this
which did not prevent sales
of go games around the world at
across the globe to soar
sign that it is not because the
machine or exceed that we change
our habits
necessarily and then with alpha go we
have learned about ourselves
especially that we are well
worse go player than we
believe we had this species
arrogant to think we were close
perfection
and finally we were very far and
we came here
one of the essential goals of the pioneers
artificial intelligence understand
ourselves through the reflection that we
tends the algorithm
so legault so extraordinary what
it’s very late it’s a game but we
can involve intelligence
artificial has no other purpose which are
of vital interest especially health
and there is no doubt that some
algorithms suffer the opinion of some
of us in the future in terms of
health is not like in the gault on
know our failures and our
weaknesses
the difficulty of making diagnoses
and all the biases that are related yes
this image at the top left
sorry if we can go back
thank you yes this picture on the top left
it’s a radio used to test
the acuity of the human diagnosis but one
did reproduce a variant of a
very well known experience in science of
commission is inserted and you see it
maybe an image of a gorilla and
believe it or not the vast majority of
experts who are asked to analyze
the image does not notice that they have a
little gorilla that was inserted into the
radio regulars that they are to detect
some kind of image and yes these are
forgotten our habits and it takes in
be aware
then there are other problems in
medicine that are considerable in
particular the fact that his door on
so many parameters so much
possible pathology that is struggling to
identify the levers the tenants and
the leads but also some
explains to you with arguments that he
must be taken seriously that the majority
of the research that is published in this
field in other sciences sciences
to the living are fake most of the
time due to lack of statistical rigor
however, we have high hopes for
does that there is a revolution that is in
medical walk because we have
access to new quantity given by
report to ourselves and of course the
possibility of sequencing
large scale genomics but also
new methods of exploration the
Clarity method developed in the United States
where an ultrasound of the future develop
in France in the wake of research
of mathias fink on the reversal of
wave
in short a lot of new data that will
enable us to perfect our
knowledge ourselves the data there
must be interpreted
sometimes it’s difficult in one of his
most important texts stephen jay
gould tell us about his emotional journey
and intellectual when he wanted in
know more about the cancer that had
touched himself after the operation the
that’s searching the literature
scientist on this subject and who
discovers that he has little hope for
have the statistics are depressing
the median survival very low and then
he thinks and thanks to his training of
scientists understands that
take into account all the data
statistics and that finally the
median is not the only indicator
statistics is important to him and
finally he survived this cancer
this reminds us of how much it is
difficult to interpret data
sophisticated that it is small
number or in large numbers as part
logical goal we messed up
extraordinary for example for the
genetics and this frustration
extraordinary that we know so well
know the genome and we have so much
to badly define in deduce the
phenotype the way we got ourselves
constituted artificial intelligence
provides a track
to solve this dilemma in a way that
is not fully satisfactory well
sure but that is impressive in
giving to learn to the algorithm
number of examples of genotype and
phenotype
in this example the researcher deviation of
sabbatini in a talk to ted us
shows how the algorithm works
how he is able for example to
restore the likely face from
of the genome here right the face
rebuilt from his own
genome left his true face with
few more elements it is true the
freedom to play on his hair system
a little bit of overweight he responds
compared to the simulation but when
even a great precision
it makes you dream it also evokes all
kinds of debates that are sure to
to ask himself who will testify before the
court of such reliability
methods when they are used in
surveys and a lot a lot
other topics and then the intelligence
artificial will also serve for the
training for the training of our
doctors our surgeons for the
better knowledge of our organs
she will serve again she is already serving to
make smart dentures that
adapts to the needs of people to
to enable people with disabilities to
discover new possibility of
pilot a mechanism a drone a notebook
by the movement of the eyes or even by the
thought among the beautiful start-ups
amount of start ups that are born on this
topic on intelligence
artificial and health in France and
elsewhere
I will mention what cardiologist
does the diagnosis
Automatic from
electrocardiograms a small
revolution in the upcoming field
or even pixel rats that won
just recently a global contest on
automatic cancer recognition
Breast from mammogram sign
if it is and the quality of the start
up french in the matter
in a subject of interest
bigger and bigger with
artificial intelligence
contact between the expert and
sector people is vital that this
from the pragmatic point of view
theoretical point of view there are some
emmanuel years when of this
mathematician stationed at stanford
one of the experts in certain fields
artificial intelligence tell us about
his ongoing collaboration with
doctors in a hospital to put in
point of new techniques analyzed
signal that can reduce a
factor 8 the time it takes us to
pass in a scanner an example of
exciting collaboration that opens up
new ways and new dreams
in the dialogue between mathematics
and other disciplines with
artificial intelligence we
let’s get to know our
own biology and just recently thanks
to recognition techniques
automatic image has been solved a
Centennial controversy over the
structure of our brain existence
of what is called the fascicle
vertical occipital an example among
many others
fascinating topics topics that
takes us to new horizons
scientists and the topics he goes
it’s also about putting music to music
better with society and
contractors
through new places for the
dialogue
new modes of collaboration like
in the collaborative project epidemic
are carried by a laboratory
of citizens innovation the benchtop and
rock labs putting together
very many skills at the service
of the fight against cancer
however, the path is paved
with obstacles
no later than yesterday at a
press conference with the minister in
digital charge
mounir mahjoubi I presented the
first conclusions the progress report
of our mission on intelligence
artificial by making the list of
many brakes on this area that remains
to solve issues related to our
economy and innovation policy it
is our training system
barriers and our organization in
terms of collection to share data
which are so important to feed
the new brake algorithms
legislative financial cultural material
all kinds of obstacles that we
will be able to defeat
I am sure at the level of France is
at the level of europe because the game is worth
the candle and because it is exciting
artificial intelligence is a subject
which allows the whole society to
dialogue and that also allows us to
we explore ourselves to discover
in our human organizations or are
forgotten us strengths and our weaknesses
and progress is a subject too
which brings us to new horizons
exciting scientists just recently
I have the work book the sni valiante one
the most original theorists of
computing it so this time on the
fact that there is going to be a strong parallel
between performances
extraordinarily strong
unexpected intelligence algorithms
artificial immersing for example
neural network techniques and
incredible performances unexpected that
showed the evolution of species with strong
organized complex organizations
structured incredibly
effective finally maybe that
evolution of algorithms and
the evolution of biology is
intimately related and maybe I
would return in this way to interest me
to my childhood loves about evolution
species so for sure it’s all about
a story to follow
more together thank you
[Applause]
”
Be First to Comment