presentation with a question on which
photo is a human pretty simple it’s the
one on the right what about now still
pretty simple it’s the one on the right
but on these two photos we can see how
fast artificial intelligence has evolved
over the past couple of years but first
we need to define what is artificial
intelligence AI is a machine that
behaves intelligently like a human it
can be by walking around or having a
conversation with somebody or it can be
already can understand its surroundings
so we can divide artificial intelligence
into two groups into eki and it’s a
strong AI and this has nothing to do
with how strong they are physically but
with the task they do weak artificial
intelligence does very simple narrow
tasks that we program it to do this can
be calculating something on a calculator
which is a very simple example and
probably everybody already use a
calculator before and programming weak
AI is a lot like baking a cake you only
can get out what you put in so very
simple and narrow tasks then we program
it to do so we’re very good at creating
weak AI and a more complicated and
advanced example of weak AI is Baxter
he’s a robot that is supposed to help
you move around objects and this can be
in a factory by picking up water bottles
and put them in into boxes or he can
help you fold your clothes or maybe even
put together IKEA chairs because you’re
too lazy to do it and we’re so good at
making weak artificial intelligence that
we can teach it how to be better than us
in some things and a great example of
this is Google’s project deep mind
alphago and alphago is a machine that
plays the game of go they probably don’t
know where what is the game of go
because it’s probably it’s only played
in South Korea mostly but to very
quickly explain what is go
you have white or black rocks and you
placed your rocks on the board and your
goal as the player is to surround as
many squares on the board with your
rocks as you can now this seems very
simple but there’s actually more
probable outcomes of a game of go than
there are atoms in a universe so you
can’t teach this AI how to play go you
need to teach it how to learn how to
play go and Google did this by saying to
the robot that it’s supposed to count
the probability of it winning on a small
piece of the board and the amazing thing
that alphago managed to do is win four
out of five games with the best human go
player so we can not only teach a AI how
to help us but you can teach it how to
do things better than us and maybe some
of the most advanced artificial
intelligences that we have now is Sophia
Sophia is a sentient AI which means that
you can have a conversation with her she
can talk with you and she can respond
how she likes and the amazing thing that
Sophia is achieved is that she’s the
first artificial intelligence to become
a citizen of a country this was a purely
symbolic move by Saudi Arabia but still
in the future when we’re going to look
back at 2017 we’re gonna say that that
was the year in which Sophia became the
first-ever citizen so not only or is
artificial intelligence helping us in
everyday lives it can also do things
better than us and it’s being it’s
becoming it’s being treated more and
more like humans and the second
artificial intelligence is Watson Watson
is a cognitive artificial intelligence
which means that he understands broad
knowledge and facts and the amazing
thing that Watson has achieved is become
the first artificial intelligence to win
the American game show called Jeopardy
and in jeopardy you don’t only answer
questions about facts and about broad
knowledge but also you need to
understand pop culture and
which have been extremely hard for
artificial intelligence to master so now
that we know what week is weaned to
answer with strongly is strong AI is
artificial intelligence that can do any
task that an adult human healthy human
can do this can be picking up objects
walking around having conversations with
people now we haven’t created yet an
artificial intelligence that could learn
anything that we do but we’re very close
to it and professionals predict that in
2040 2050 or maybe even sooner we’re
going to have an artificial intelligence
that can do everything that we do right
now so we need to talk about this now
because artificial intelligence is
coming and not only is revealing to be
an artificial intelligence that can do
everything that you can do it may also
be conscious now we need to explain what
is consciousness
at first so nobody actually knows
exactly what is consciousness and that’s
the funny thing because you can’t put it
under a microscope it’s not a solid
state like an atom would be but we know
how we feel when we’re conscious and how
our unconscious and consciousness is
knowing that you exist and the things
around you exist so not only would this
robot it wouldn’t be also able to do
everything that you do but who also
understand the world around it and know
to the biggest extent of its knowledge
that it exists so now we need to ask
ourselves why do we have rights in order
to answer the questions those artificial
intelligence deserve rights because we
need to ask ourselves how are we going
to treat artificial intelligence that is
just as good as us well why do we have
rights
humans have rights because we believe in
human exceptionalism and human
exceptionalism is the belief that humans
are smarter than any other creature in
the world then monkeys then flies were
smarter so we deserve more rights more
privilege and the answer the question
now
this AI deserve rights well we can see
if it can do anything that we do and if
it’s as smart as this but how about Khan
how do we know if a robot is as
conscious as we are meet Alan Turing’s
Turing test and Turing test is a test in
which you communicate by writing
messages with a robot and a human and at
the end of the test after you ask
questions to the robot and the human and
they can respond by giving you a
question or lying you need to determine
if the person you were talking with was
a human and which one was their
artificial intelligence so a robot would
pass this test if it would be able if
you were not able to distinguish between
it and the human so there’s been a lot
of talking about this artificial
intelligence that would be just as smart
as us and it would be conscious of its
surroundings and the European Union took
notice and I present to you maybe some
of the most important laws that they
have created right now for artificial
intelligence and the first one says that
a robot may not harm humanity or by
inaction allow humanity humanity to be
harmed and I think that’s pretty
illogical and straightforward and
everybody can agree with that law and
the second law a robot must obey laws
and orders given it by humans and this
law is great with dealing with
artificial intelligence that we have now
like Baxter Sophia or Watson or even a
calculator right
but with robots that we’re going to have
in the future in the near future we
can’t treat them like slaves because I
think everybody can only agree that
slavery wasn’t a good idea
and we shouldn’t repeat it so I have
this very bright future in my mind
pretty utopian and very distant but I
believe that we should treat artificial
intelligence as a smart machine a smart
machine that can help us that we should
treat with respect and we should try to
understand it and try to give it a bit
of room for itself to develop thank you
very much this is my idea worth
spreading
[Applause]
[Music] [Applause]