at least one certainty
the human sentient being the most
evolved of the solar system
there is no living being able
to create cultural economic structures
and social as complex as the
human beings
this certainty is so rooted in
we who do not think about it and think of it
be out substantially from one
evolutionary travel of expenses and therefore
have reached some goal
cognitive no longer reachable from
no living being
we are sure it is so
in fact we ourselves are in the
helping to create is an intelligence
the artificial one that could
pass us late until the 50s
many researchers are researchers
dedicated to artificial intelligence
to create machines that could
solve very complex problems and
so he was not able to solve
in other ways these intelligences yes
have been based on models since then
statistical mathematician able to
provide the computational elements in
a specific domain
so that the machine can
basically to take that domain
make categorizations and give some
forecast compliance with the scenarios that
we do not introduce him now does not sound much
intelligent as a definition, however
it’s what they basically do
machines at the moment what
he attributes intelligence to us
basically we see when
we see intelligent behavior
we think that people actually do
in reality it is all excellence
artificial at this time is based
basically on capacity
then computational these objects
they have enormous computational capacity
this kind of artificial intelligence
it is called artificial intelligence
weak in contrast to the
concept instead of intelligence
artificial strong where the theory is theorized
possibility of doing
think of the machines and give them
self-awareness from their self-awareness e
let them reason in a similar way
to the human being
these two concepts have a nuance
let’s say various nuances of intelligence
that we can identify the first are
reactive machines to keep active
they are the first stage of these kind of
intelligence when they can do
they can analyze a domain extract
of a range of possibilities e
choose the best one for a purpose
specific here you see in the image deep
blue that perhaps one of the first and most
famous and artificial intelligences of pm
who defeated the world champion of
kasparov chess already in 96 when
kasparov only a few years before had
beat 32 computers at the same time
up in matches on ss chess states
so what happens begins to
increase computing capacity and in
only last year google pulls
out to the fado that is an agreement
artificial that plays go go is a
game much more complex than chess
and it was thought he could not be beaten
a human being instead last year there
if from the sample the Korean is
been beaten by this car it is
exit another just a few
month ago that is called to the phage 0 that for
the first time he can learn from
only then the phage was based on
the human experience to beat the
alpha humans with zero is based on the
his only experience therefore he played in
three days something like 300 million
leave and now shred any
previous version of alpha 2 and then
this here is the typical artificial that
we have at this moment the next ones
they are those that
they are called intelligences
artificial memory limited thing
they make these intelligences start to
perceive the world around them
therefore they are equipped with sensors with the
which they manage to create one
fairly reliable representation
of the space of time they have a memory
imitated so they can not maintain
like the brain
these experiences therefore can not
use their experience but they can
however interacting much more
sophisticated this is the example here is
a self-driving machine and that
that is approaching at this moment of
more to this kind of intelligence
then we have a type of which not
we still have examples that are
basically machine that yes
they refer to the theory of the mind
so we add a feature
I perceive the world I understand space time
and here I also understand that I have agents
intelligent that therefore are not gods
of the physical objects that they then have
of the thoughts of emotions and of the
motivations and I take the acisco of
therefore I begin to interact
example with the human being in contexts
different social or cultural emotions
so I understand when the human being does not
he wants to be annoyed rather than instar
they have happy resources at the last stage
of artificial intelligence is that
less and official agreements
self-conscious here we are very far from
getting to this point is so much that it does not
we understand well less the knowledge and
human consciousness and what it means
self-consciousness so here is a tabula
rasa on which on what system
focusing on refining
reactive machines and the evolution of
that memory is limited I’m not alone
artificial intelligences from the laboratory
or by experimentation play
let’s talk about intelligent that you
you already have summer already using you have a
portal to these intelligences
artificial that to your smartphone
access substantially to assistants
virtual as you remember google now
now, however, ignited by these
type of virtual assistants is based
basically on machine algorithms
learning with more than agrifids like
let’s say analysis
behavioral to customize the
answer to your needs
here too it is essentially based
on the computational capacity and more
on the data very important because
why these special arrangements like the
most of the future will be based on
data produced by you and you are the
fuel of this intelligence is
to perfect both to make them evolve
what happens so we have a lot of
data that are ours that are put to
arrangement of machines and
controlled by very specific subjects
that is, in particular this big moment
company of the information technology industry
with a proactive step of
self-regulation many of these
hi tech industries said ok we do
a partnership we demonstrate that
the intelligence of life is a beautiful thing
positive we can do a lot of
things that we list in the guidelines
that we will follow so that we are
sure that these intelligences
Artificial will do what it takes
in this way we say we try not to
sterilize a market that is nascent
it is very promising a little bit let’s say
against this opinion is this gentleman
that I quote because it says what
interesting is that only product are a
contact with artificial intelligence
truly avant-garde and I think that the
people should really be
worried about keeps on playing
the alarm bell but until when
people will not see robots around
on the streets to kill people not
they will know how to react because everything
It appears so vague and still
artificial intelligence is rare
case in which we need to be
proactive in regulation instead
that reactivate I think that at the moment in
which we will be reactive in the argumentation
of official sla will be too late
this is what the on mask says in a
male for the few who do not
they knew he is a visionary
contemporary businessman the man who is
behind extraordinary companies like
space ex that was for the first time
in history he was able to
make reusable rockets then
opening the exploration substantially
spatial and other colonization
planets for being for mankind
or
founded companies that can
focuses on making all the energy
of the world as sustainable therefore
making central central violets
solar energy clean and realizing ad
example the first electric machines a
autonomous guide then one that we say of
forecasting technologies are intended
this sentence told her of his
little room called said before the
American government in a hearing
public just to raise awareness and
for the first time let’s say it’s rare for
the American entrepreneur claimed that there
he wants a control that comes
absolutely before innovation
of course, but you can not imagine them
terminators that revolve the being
human there is absolutely no need of
get up to why because the
point is not that the point is that
artificial intelligence will not be that
we are smart to do things
extraordinary and intelligence that
we mean human
the point is that here the intelligences that
they are concentrating in the hands of
very few subjects and then become
a weapon an absolutely powerful weapon e
towards which he has overcome a certain
we will no longer be able to rebel but
as basically made jazz is
official there are so many ways to
do it but the most efficient ones
they basically provide for interaction
of an infinity of systems systems that
they interact with each other and work
data and give results then
we basically imagine sensors
of actuator processing
now this type of architecture and has
need a lot of features
we say techniques I will not list you but
basically we talk about
reliability of data certification
bandwidth speed there is a whole series of
things we could actually do
we can find something that
we already have the internet
now the internet is born as a network
unfortunately distributed over time for
questions of practices first and let’s say
opportunistic then slowly becomes
an increasingly centralized IT network
therefore where of the subjects that generate e
we say not monitor and manage
practically all the world traffic
with the increase in quality
actually the services you are interested in
offers
is a spasmodic increase in users
what happens to this point the typical technicians
doctor cabbage we can not do it
the infrastructure does not hold up
what is done is centralized even more
so what you call now
cloud is actually gigantic
data processing centers that they manage
practically almost all crimes
world by centralizing a power
incredible now fortunately in
parallel to these this kind of
trends are born the technologies that we
we love peer to peer and therefore peer
therefore and substantial are technologies that
allow not to go from servers
central to be able to interact
typical example is that of the file
sharing then file sharing
rather than rather messaging
that video streaming now this type
trends that were previously limited are
substantially to a moral choice
rather than a philosophy of privacy
right in privacy transparency
information and anything else
now it is becoming indispensable as
as an approach we say architectural
because the number of connected devices
we had also said in other talks
increasing exponentially here we speak
tens of billions of devices that yes
will enter this network as users and
so they will want to interact then
that infrastructure, however, not
hold up
so we are going towards delle
decentralized architectures thing
means means going to work on
an archive of an architecture the first
the most the most famous one too
tool is more powerful is and slave
there is substantially rendered knowledge
It has been made famous in recent years
in particular from the bitcoins which are one
crypt that basically has
let’s say so a little invented the
burning of its purposes
what can you do with a brussels for i
non-technicians were united imagined
in a book a ledger of
information in which each actor can
writing something is not necessary
the important we say central point that
you do not need a central entity for
govern this type of network here
information can be written
validated and confirmed and without the need for
the central authority this yes is
this is done through mechanisms of
algorithms consent mechanisms
distributed is very important the
consensus concept precisely because yes
make knots accountable and avoid
go to rely on individual subjects
now this type of architecture e
provides the ability to write both
for example, I give the unknown
someone then a simple transaction
financial but above all they can
also create small programs
the difference with traditional g that here
the program
I know exactly what it will do in the face of
a certain event so I wanted to
very simple you can write a
program to which he prizes a prize
insurance against the occurrence of
a sinister ok this thing now
It is made by 300 subjects and that is
subject to change here I do not know that
it would not be
in addition this type of networks is based
basically on a cryptography ai
maximum current levels so everything
what you do now your card
of credit
the accessory is the bank sites
rather than online messaging
these technologies do it with a degree
even greater security and are already
some ready to face the saying
the problem of cryptography
quantum and that is
the ability to resist usually
task tremons that are next
evolutionary stage we say of capacity
computer computational then
we have this kind of network what thing
we can therefore do at this point
we can imagine the creation of one
official intelligence that goes
basically to use of the
programs are smart contracts
transparent and that constrain what
it is an artificial type intelligence can
make of defining its boundaries of
its foundation so that the intelligence
artificial can evolve but remain
under our control
what is needed to do this thing you need
a democratic structure and the
participation of all actors a
this kind of intelligence
so what do you ask what we have to do
do to make this kind of
intelligence we must surely
need technicians to create tools
adequate to provide the ability to
expressing the will in a manner
democratic then there should not be a
subject that can be bound to express
this will
we can make the interface bad
indeed we can make the networks fast
we can do a lot of things but then
you need someone to make it work
this smart motor cap
this is you, you are the ones who
they will give the data and they will give the indications
to this intelligence
so what is needed is the commitment to
be aware of one’s own
role while now you are therefore passive
Tomorrow should have this
awareness
if we want to continue to say
declare ourselves the living species more
evolved of the solar system no more
this time as individual subjects but
as a collective sentient entity
thank you so much
[Applause]
”
Be First to Comment