[Music]
although I love dogs very much I will deal with one
subject of animals in general but
especially dogs will treat one
argument that somehow yes
reconnect a treaty first by a
previously by a colleague is
the topic of intelligence
But I just said I ‘m one first
scientist I ‘m an IT specialist
I am part of that category of
scientists who were not even that much
considered intelligent so there
try even to give the definition of
intelligence I do not define intelligence
but I will follow a very brief path in
which I will tell you what we see as
we see precisely this question of
comparison between human intelligence
artificial intelligence of course
I will pose a problem
I will define a problem finally I will close with
an example that I learned a few
weeks ago a few months ago and that me
really like it so let’s start
let’s start with the fact that some scientists
much better than me some time ago
they started talking about singularity
meaning by this term
that exact moment in which
the intelligence of the machines would have
exceeded the intelligence of men
once defined this concept
concept of singularity they are
unleashed by scientists to imagine
when this moment would arrive
values I want to try to give you some
information to leave then to you
decide if the singularity is close
near far far away
impossible
the information I give you talk about
machines in the way that i
informatics talk about machines i
informatics talk about machines in
terms of algorithms what is a
algorithm when I started studying
these things thirty years ago and for a long time
time still an algorithm was one thing
simple it sounds like a song but it was like that
an algorithm was a simple thing a
algorithm from simply a recipe
to process data
if he has then b otherwise c
if it rains, take the umbrella otherwise
leave the house
so simple recipes to deal with
data in a manner with an aura above
impartiality of neutrality it has given
in all of us members of great humanity
trust in algorithms
we have trusted the algorithms and
the algorithms have entered into ours
lives I’m not to mention the millions of
applications in which the algorithms
in fact they play an important role e
they accompany us but from medicine to
finance
we are at the transport science
entrusted so much that we have so many
in many contexts entrusted to them the
the task of choosing does not seem to be the case
exaggerated what I’m saying
because your experience yours
daily human experience for example
when you go to a doctor without
to offend the doctors who are here is
an algorithmic experience asks you to
undergo examinations that return gods
number data based on feedback
of those numbers compared to certain
parameters makes a decision
he actually makes the decision
it takes the algorithm so we are there
in love we trusted the
algorithms and we have entrusted them
many of the choices that we humanity
we would have done so far as I said little
bad if a problem we gave up
choose on things on which precisely there
we are trusted but all in all no
problem but because no problem
because those algorithms as he said the
colleague who preceded me
they were nothing but knowledge
intelligence if we want to say that
the one who had written the algorithm
the one who had conceived the one who did it
had planned had entered into
that algorithm then an algorithm was
somehow simple even though
knowledge transmitted by a human being
so why not trust us if we trust
of the human we trust the algorithm that
he has designed human to date everything to
place
but things change sometimes
also very quickly
and they change very often without that
we notice it or with humanity one
a little distracted change current point
that now in my sector we talk about
ag
agi means artificial general
intelligence I do not have the claim then
make a lesson on what it means
this strange thing but I’ll try to
explain it with very few words
succinctly starting from what
surely you understood to be one
traditional algorithm ie an object
first we said absolutely
hobby and obedient absolutely
foreseeable absolutely photocopy
of the knowledge that the human had them
now transmitted these new ones
algorithms are a new thing not me
ask why as from the point of view
technological technician this was
possible however what happens that
this new class of algorithms is not
more knowledge only but it is knowledge
experience units ie these algorithms
these machines are able to have
experiences are subjected can
be subjected daily to
tens of thousands of millions
billions of example cases from which
they learn things that were not initially
codified in the initial knowledge a
modern algorithm therefore an algorithm of
artificial intelligence those things that
they call deep learning machine learning
neural networks a modern algorithm I was saying
it no longer responds to the simple equation
algorithm equal to knowledge of the human
but it responds to a new kind of equation
that is the same algorithm plus knowledge
experience experience conducted in a
broad context like reality like the
world in which we live well if you have
understood the difference if you understood the
definition probably started too
to understand the problem in a moment beer
i’ll tell you i’ll explain the problem in
terms I think understandable but as
university professor I remain minimal
20 seconds I want to do it and I want it
to do by mentioning the scolding I took
this summer by an old colleague
American that when it was reflected on
these things he told me is Marco as al
soul calls me young but I am not
younger marco as usual you
young people we teach you young but you
if it seems to me that you understand but forget
you have forgotten when you
ask this theme the new algorithms are
something different from the old ones
incorporate the ability to do things that i
first they did not and start to
worry with many others in the
world
have you ever forgotten the first thing you do
they teach when we taught you
cybernetics
the first law of cybernetics yes
called ashby law says that if you
do you want to control a system of
a certain complexity you must have a
layout a controlling system
as complex as the system that
you want to check and then he would tell me
but sorry, you wonder now
product and to be producing systems
that are more and more algorithms
complexes that always make decisions
more complex in more and more contexts
vast ones whose consequences whose
implications can very often be
unpredictable but because you are wonderful
you have created are creating we are
creating a network a sort of reality
parallel with the same controller
complex compared to what we wanted
to check
here is therefore the theme here is the
problem if a question arises of
comparison between the intelligence of these
new machines and human intelligence yes
poses exactly in these terms
these new classes of algorithms
learning from the experience they do
every day and not coding then
more alone not having only more inside the
knowledge of the engineer
of the designer who opens them
drama the may have behaviors
unpredictable otherwise you do not
would explain the woman who going in
bicycle
about two months ago in a rice road
of Arizona in full sun has been taken
below by an autonomous driving machine of
unpredictable huber lim predictability me
approaching the conclusion of this
intervention
so imagining that maybe if there is one
answer to this problem is one
response that revises in some way
the alliance I speak precisely in these
almost biblical terms
the alliance between men and machines
for men and algorithms from mine
point of view from computer scientist
for example I believe that if there is a new one
I will amaze you at the frontier of technology
so, but this new frontier e
ethics I see no more fervid period anymore
most important fertile in the future for
a discipline that is not for example
philosophy because we will have to re-discuss it
together what it means to blame
what does it mean wrong?
what does intentionality mean?
what does negligence mean?
I close with an example of a beautiful one
alliance between men and machines going down
a little from the empyrean or cloudy on which I have
lived so far to an example instead
real war of Syria between 56 million
of refugees 500 thousand dead
the 500 thousand dead counted that I mentioned
they died for the fights between the
rebel factions rebel and factions of
assad supported by the Russian bombing
in this war in that war all
they have their hands stained with blood
even the United States has killed around
6,000 isis rebels all have them
hands dirty with blood and
if you speak or if you read articles by
people living in that hell that
they lived in that hell narrate
that one of the scariest things like
always for wars and if only for
modern wars are bombings
[Music]
at this moment in particular there is the
idlib area which is a small enclave
in which almost all the refugees have taken refuge
rebels and being bombed
heavily from the militias as a list
supported by the Russians
it is estimated that around 20 percent of
those who are in this club have
in reality dealing with the rebellion
eighty percent are normal people
like me like you and therefore suffer
bombing in some way
unfairly compared to their own
condition of being normal good three
activists take shazam shazam is a
very nice I’m sure many of them
you know him a very nice one
application of artificial intelligence
which therefore has inside all those
strange algorithms that learn and speak
more and more of which I talked to you before and
that recognizes a song
if you want to enter a pub there is one
song that you do not hear thrown shazam
and shazam tells you what song it is
treating these three to three well
activists have modified shazam have
built a series of sensors that have
distributed in the woods and meadows in the
fields where possible in the area of
slip gathering no more noise of the
songs but the noise of war the
noise of the bombers these noises
they become data that are collected by
an artificial intelligence that
using sophisticated formulas of
physics type the doppler effect hand a
hand that collect data always learn
better for a bomber than it is
starting is going into a certain one
position where he will bomb who he is
connects to this application that yes
call centers
can do with telegram can do it with
whatsapp can do it with twitter
right now at idlib at ten minutes
to escape before a bombing
these ten minutes are worth
the alliance between men and machines
[Applause]
although I love dogs very much I will deal with one
subject of animals in general but
especially dogs will treat one
argument that somehow yes
reconnect a treaty first by a
previously by a colleague is
the topic of intelligence
But I just said I ‘m one first
scientist I ‘m an IT specialist
I am part of that category of
scientists who were not even that much
considered intelligent so there
try even to give the definition of
intelligence I do not define intelligence
but I will follow a very brief path in
which I will tell you what we see as
we see precisely this question of
comparison between human intelligence
artificial intelligence of course
I will pose a problem
I will define a problem finally I will close with
an example that I learned a few
weeks ago a few months ago and that me
really like it so let’s start
let’s start with the fact that some scientists
much better than me some time ago
they started talking about singularity
meaning by this term
that exact moment in which
the intelligence of the machines would have
exceeded the intelligence of men
once defined this concept
concept of singularity they are
unleashed by scientists to imagine
when this moment would arrive
values I want to try to give you some
information to leave then to you
decide if the singularity is close
near far far away
impossible
the information I give you talk about
machines in the way that i
informatics talk about machines i
informatics talk about machines in
terms of algorithms what is a
algorithm when I started studying
these things thirty years ago and for a long time
time still an algorithm was one thing
simple it sounds like a song but it was like that
an algorithm was a simple thing a
algorithm from simply a recipe
to process data
if he has then b otherwise c
if it rains, take the umbrella otherwise
leave the house
so simple recipes to deal with
data in a manner with an aura above
impartiality of neutrality it has given
in all of us members of great humanity
trust in algorithms
we have trusted the algorithms and
the algorithms have entered into ours
lives I’m not to mention the millions of
applications in which the algorithms
in fact they play an important role e
they accompany us but from medicine to
finance
we are at the transport science
entrusted so much that we have so many
in many contexts entrusted to them the
the task of choosing does not seem to be the case
exaggerated what I’m saying
because your experience yours
daily human experience for example
when you go to a doctor without
to offend the doctors who are here is
an algorithmic experience asks you to
undergo examinations that return gods
number data based on feedback
of those numbers compared to certain
parameters makes a decision
he actually makes the decision
it takes the algorithm so we are there
in love we trusted the
algorithms and we have entrusted them
many of the choices that we humanity
we would have done so far as I said little
bad if a problem we gave up
choose on things on which precisely there
we are trusted but all in all no
problem but because no problem
because those algorithms as he said the
colleague who preceded me
they were nothing but knowledge
intelligence if we want to say that
the one who had written the algorithm
the one who had conceived the one who did it
had planned had entered into
that algorithm then an algorithm was
somehow simple even though
knowledge transmitted by a human being
so why not trust us if we trust
of the human we trust the algorithm that
he has designed human to date everything to
place
but things change sometimes
also very quickly
and they change very often without that
we notice it or with humanity one
a little distracted change current point
that now in my sector we talk about
ag
agi means artificial general
intelligence I do not have the claim then
make a lesson on what it means
this strange thing but I’ll try to
explain it with very few words
succinctly starting from what
surely you understood to be one
traditional algorithm ie an object
first we said absolutely
hobby and obedient absolutely
foreseeable absolutely photocopy
of the knowledge that the human had them
now transmitted these new ones
algorithms are a new thing not me
ask why as from the point of view
technological technician this was
possible however what happens that
this new class of algorithms is not
more knowledge only but it is knowledge
experience units ie these algorithms
these machines are able to have
experiences are subjected can
be subjected daily to
tens of thousands of millions
billions of example cases from which
they learn things that were not initially
codified in the initial knowledge a
modern algorithm therefore an algorithm of
artificial intelligence those things that
they call deep learning machine learning
neural networks a modern algorithm I was saying
it no longer responds to the simple equation
algorithm equal to knowledge of the human
but it responds to a new kind of equation
that is the same algorithm plus knowledge
experience experience conducted in a
broad context like reality like the
world in which we live well if you have
understood the difference if you understood the
definition probably started too
to understand the problem in a moment beer
i’ll tell you i’ll explain the problem in
terms I think understandable but as
university professor I remain minimal
20 seconds I want to do it and I want it
to do by mentioning the scolding I took
this summer by an old colleague
American that when it was reflected on
these things he told me is Marco as al
soul calls me young but I am not
younger marco as usual you
young people we teach you young but you
if it seems to me that you understand but forget
you have forgotten when you
ask this theme the new algorithms are
something different from the old ones
incorporate the ability to do things that i
first they did not and start to
worry with many others in the
world
have you ever forgotten the first thing you do
they teach when we taught you
cybernetics
the first law of cybernetics yes
called ashby law says that if you
do you want to control a system of
a certain complexity you must have a
layout a controlling system
as complex as the system that
you want to check and then he would tell me
but sorry, you wonder now
product and to be producing systems
that are more and more algorithms
complexes that always make decisions
more complex in more and more contexts
vast ones whose consequences whose
implications can very often be
unpredictable but because you are wonderful
you have created are creating we are
creating a network a sort of reality
parallel with the same controller
complex compared to what we wanted
to check
here is therefore the theme here is the
problem if a question arises of
comparison between the intelligence of these
new machines and human intelligence yes
poses exactly in these terms
these new classes of algorithms
learning from the experience they do
every day and not coding then
more alone not having only more inside the
knowledge of the engineer
of the designer who opens them
drama the may have behaviors
unpredictable otherwise you do not
would explain the woman who going in
bicycle
about two months ago in a rice road
of Arizona in full sun has been taken
below by an autonomous driving machine of
unpredictable huber lim predictability me
approaching the conclusion of this
intervention
so imagining that maybe if there is one
answer to this problem is one
response that revises in some way
the alliance I speak precisely in these
almost biblical terms
the alliance between men and machines
for men and algorithms from mine
point of view from computer scientist
for example I believe that if there is a new one
I will amaze you at the frontier of technology
so, but this new frontier e
ethics I see no more fervid period anymore
most important fertile in the future for
a discipline that is not for example
philosophy because we will have to re-discuss it
together what it means to blame
what does it mean wrong?
what does intentionality mean?
what does negligence mean?
I close with an example of a beautiful one
alliance between men and machines going down
a little from the empyrean or cloudy on which I have
lived so far to an example instead
real war of Syria between 56 million
of refugees 500 thousand dead
the 500 thousand dead counted that I mentioned
they died for the fights between the
rebel factions rebel and factions of
assad supported by the Russian bombing
in this war in that war all
they have their hands stained with blood
even the United States has killed around
6,000 isis rebels all have them
hands dirty with blood and
if you speak or if you read articles by
people living in that hell that
they lived in that hell narrate
that one of the scariest things like
always for wars and if only for
modern wars are bombings
[Music]
at this moment in particular there is the
idlib area which is a small enclave
in which almost all the refugees have taken refuge
rebels and being bombed
heavily from the militias as a list
supported by the Russians
it is estimated that around 20 percent of
those who are in this club have
in reality dealing with the rebellion
eighty percent are normal people
like me like you and therefore suffer
bombing in some way
unfairly compared to their own
condition of being normal good three
activists take shazam shazam is a
very nice I’m sure many of them
you know him a very nice one
application of artificial intelligence
which therefore has inside all those
strange algorithms that learn and speak
more and more of which I talked to you before and
that recognizes a song
if you want to enter a pub there is one
song that you do not hear thrown shazam
and shazam tells you what song it is
treating these three to three well
activists have modified shazam have
built a series of sensors that have
distributed in the woods and meadows in the
fields where possible in the area of
slip gathering no more noise of the
songs but the noise of war the
noise of the bombers these noises
they become data that are collected by
an artificial intelligence that
using sophisticated formulas of
physics type the doppler effect hand a
hand that collect data always learn
better for a bomber than it is
starting is going into a certain one
position where he will bomb who he is
connects to this application that yes
call centers
can do with telegram can do it with
whatsapp can do it with twitter
right now at idlib at ten minutes
to escape before a bombing
these ten minutes are worth
the alliance between men and machines
[Applause]
Please follow and like us: