Press "Enter" to skip to content

Artificial Horizons | Martin Ciupa | TEDxTanglinTrustSchool


so am I am i able to be heard
good well just as we’ve seen I don’t
know if it’s really accurate to say that
I am from artificial intelligence it
doesn’t really sound right does it I am
a citizen of the UK but I’ve lived a
major part of my life in Singapore and
in Sydney and in Seoul Korea and I’m now
living in New York State
so yeah I’m kind of mixed up I suppose
anyway what I’m going to be talking
about is artificial intelligence and I’d
be particularly looking at it from a
point of view of the artificial horizon
which is which we are approaching I’m
going to also be talking about robotics
and I’m going to be placing these in the
context of a dialogue which I’ve called
the prophets of doom and the doom of
prophets were actually the doom of
prophets being put first in the business
of artificial intelligence and in
industrial automation and robotics I
should say that I’m the chief technology
officer of a robotics company in the
States so that gives you a little bit of
perspective of where I’m coming from
now AI is booming we’re talking about
300% growth in the last year we’re
talking about Price Waterhouse Cooper
claiming that by 2040 to be something
like what’s the number fifteen point
seven trillion dollars created by this
these businesses just give you a
perspective of that huge number that’s
more than double the amount of gold ever
mined by humans in the history of the
world it’s a huge number so it’s a huge
hope it’s a huge utopian goal and
likewise in industrial robotics and
automation the concept is is that goods
of high quality can be produced at such
low cost that we will have this concept
of the abundance economy but what’s
going to have
in terms of jobs what’s going to happen
in terms of our culture with the
introduction of these technologies these
are real issues and this is the main
part of the discussion likewise is not
just about factories in industrial
robotics robots are coming into our
homes artificial intelligence is coming
into our homes there are issues here
that I’m going to explore so let’s talk
first of all about the prospect of super
intelligence a kind of AI that we
actually call AG AI artificial
generically intelligence now what I
would like to draw your attention to it
is that in the sixties experts were
saying it’s 30 years from now in the 90s
and in the 90s it was in 2020 30 years
and at 2014 when this data was collected
2015/2016 they’re saying it’s around or
the major cluster is around 2040 – more
or less and that’s suggesting that it’s
less than 30 years so maybe there has
been progress in as much as it’s
becoming now below the 30 year horizon
and let us say between 15 and 40 years
is a good mark because there’s a lot of
diversity in this expert opinion part of
the situation that we find ourselves in
is well described by a paradigm of
Industrial Revolution we’re familiar
from our history books about the first
Industrial Revolution about steam and
mechanical engines and we are familiar
perhaps with the law decree billion that
occurred in textiles I actually come
from a town fifteen miles away from
where there was a major Luddite
rebellion actions in helm Shaw in
Lancashire and Industry 2.0 the end of
the 19th century beginning of the 20th
century where the revolution or the
technology driver was an
in the 1950s and 60s the revolution was
in VLSI and microprocessors and digital
and right now we are in a new industrial
revolution of the Internet of the
virtual of the cloud and there’s a new
revolution coming very quickly on its
heels
artificial intelligence something which
is radically capable of changing the
world in which we live in and the
practices that we take for granted and
the culture that we have and robotics
has played a part in all of these there
was the jacquard loom in textile mills
in the first Industrial Revolution and
actually robotics can be seen in the
electrical you know the production lines
production lines of the Ford Motor
Company Henry Ford is a kind of
industrial automation and again with
robotic arms in the digital era and now
in the connected era we also have the
robots some of these robots that we find
in our homes so I think I should be
pointing over here rather than so
basically what do we have here but I’m
talking about now that was the utopian
dream but there is a dystopian
undercurrent here and in particular
there is the concept that you’ll hear
about the existential threat of
artificial generic intelligence super
intelligence will these systems be our
overlords will they be the ones in
charge because they can operate well
beyond our capabilities and
intelligences in complex situations that
we’re incapable of managing
there is also the concern that
industrial automation in particular is
going to savage our job prospects not
only in terms of manual labor in an
industrial automation setting but
through AI in the white collar industry
Legal industries education sectors
medical and health sectors and the
military are all open for AI to have
massive impact and it’s more than just
jobs as I will be pointing out are these
hopes and fears being overhyped I’m
going to show you some figures or
diagrams I should say the first one here
talks about the media frenzy it’s been
fairly constant the blue line represents
positive media coverage of AI the orange
line red line depicts negative media
attention and currently up to about 2015
the ratio was three to warn in terms of
positive now it’s more like one to one
there is a significant rise in the fears
that AI and industrial automation and
robotics can deliver and that’s rising
dramatically so fast that we can talk
about the expectations of a hype cycle
are we in this situation were the
expectations of industrial automation
and robotics is so high that it can’t go
any further it has to come down as we
deliver production of these systems
their expectations perhaps will not be
met and then though if they really do
have value if they really can make
contribution to our society they will
rise up to a more realistic expectation
this hype cycle has been seen for many
technology innovations in our society
and right now at the top of expectations
are many of those key technologies which
relate to AI so is it type AI perhaps
and robotics perhaps more than any other
technology is very well depicted in the
science-fiction stories that you may
have all grown up with there’s one here
actually from metropolis a silent
black-and-white movie made in perhaps
the 20s I accompanied me
but then we can see many of these
pictures of some of our favorite movies
whether it’s Blade Runner or Westworld
or iRobot or the Terminator and other
such programs and these often are
cautionary tales there are stories about
things that can go wrong
they often feature rebellion but one of
the issues I’m going to put to you is
that there may well be a rebellion but
it may not be by the robots it may be
through and for us okay so I’m now going
to break up my talking to plot the
prophets of doom and the doom of
prophets being put first let’s talk
about existential risk now two
commentators I personally admire very
much for their further contribution and
capabilities in sight one is Elon Musk
one is professor Stephen Hawking yet
they’ve been very negative in terms of
their fears of AI and industrial
automation Elon Musk very recently this
week from the Rolling Stone article it’s
been talking about the chances of AI
being safe has been less than five to
ten percent I mean that’s basically
saying that AI will bring us to do
and likewise Stephen Hawking echoes some
of those sentiment though he also
mentions in the quotation here that our
economies may fail also as a consequence
of some of the job losses that will
occur through industrial automation so
just very quickly so that I keep the
time.however are those fears really
solid and backed up there are many
recent failures if you choose to look
for them for every success story there
are dozens of failures behind them
making the AI success stories is really
hard we’re not in a position where we
roll out AI systems very very quickly
they have to be tuned they have to be
nursed into existence there are known
technical problems we struggle to
understand issues like consciousness
such as free will such as
self-determination these are
philosophical concepts and we are trying
to in terms of AGI pitch those in
technological terms it’s not so
straightforward we don’t really have an
idea what those terms mean
so trying to build technology behind
things that we don’t really understand
is a genuine issue this is the mysterion
objection which is saying as I’ve said
we don’t understand these terms it may
be beyond us to build these systems the
expectation from experts let’s remind
ourselves is fifteen to thirty forty
years from now there is time to fix some
of the issues my last point is that we
can potentially fix some of the AGI
issues we can put a firewall of
connectivity between the AGI which we
may well be asking for advice for
helpers with global climate change
scenarios for example and put those such
that they don’t control things beyond
the firewall they don’t even get access
to the Internet beyond the firewall
unless it’s through a narrow
well-controlled dialogue we can actually
fix the safety issue if we have 15 to 20
years to 30 years to do that
and here’s a big point for most of the
applications AGI is totally not needed
if you’re having a robot in your home
doing your washing up doing your laundry
cleaning out the cat litter does it
really need to be conscious if it was it
might not be very happy to be doing
those jobs what we want from these
robots is to be able to do those jobs
without being happy or sad and emotion
and consciousness are not part of
so moving on now to the Deemer profit
scenario well yes
industrial automation will reduce the
amount of robotic jobs they’re out there
but just keep in mind what I’m saying
humans doing robotics jobs is not a
humane use of humanity if I articulated
that well what we really should be
aspiring to do is to assign robots to
robotic jobs and humans to humane jobs
and it’s my belief that there is an
infinite opportunity to find humane jobs
there is always need for humans to help
other humans in one way or another now
we may have to adjust our economies to
be able to support that and that’s part
of the discussion here basically we may
have to change our source your economic
models we may have to introduce concepts
such as universal basic income a whole
TED talk and more could be devoted to
that point so I’ll just leave that there
highly immersive AI gaming systems are
going to provide us with entertainment
at a level never seen before and we will
have the service robots and sex robots
available to us and the impact of those
is going to be massive there are
problems in trusting some of the deep
learning systems that we have there
black boxes we train them but we can’t
understand how they’ve arrived at their
decisions again this is a topic that a
whole TED talk could be devoted to so
I’ll move on
there are catastrophic problems in case
there are is hacking that are in case
there are bugs within the system and if
these systems have been around for a few
years the humans that they’ve replaced
their skills are not necessarily now
available after a catastrophe
catastrophe so the point is of human
atrophy we can’t roll back if we’ve been
working in mission critical systems with
AI systems the possibility is that if
they crash
it’s going to be difficult to roll back
these are major issues now they can be
resolved but they have to be planned to
be resolved
we can’t just blunder or we can I mean
humankind has had a history of
blundering into technological issues but
we have to change that mentality is my
point here otherwise we are ceding moral
authority to machines that perhaps we
shouldn’t be and we are preparing
perhaps for a rebellion but it’s not a
robot rebellion it’s a human rebellion
so I’m concluding here now trying to
keep within my summer time slot and
basically my point is don’t worry so
much about the AGI issue for
intelligence issues we can fix those
problems we don’t really need AGI in the
general community it’s more hype caused
by fear than a real risk in my opinion
but we do need to resolve and take
action today on the narrow AI
applications whether they be in health
government military legal and these can
be augmented intelligence they’re not
necessarily replacing teachers replacing
lawyers but maybe their paralegals and
their paramedics so that the third point
is is that the economic value that
massive trillion-dollar multiple
trillion dollar impact will be there for
us and furthermore that they can help us
with some of the genuine existential
issues that we have and issues of
bringing services to areas in which they
traditionally haven’t been available
bringing health services legal services
medical as I say education services to
communities which don’t necessarily have
at this time the luxury of of those in
less well developed countries I’m
finishing up as I say I don’t really
have time to read through this massive
quotation of Professor Daniel Dennett I
do hope that you look it up I understand
you
can get a copy of my slides from the
organizers basically echoes my opinion
he calls the concept of artificial
general intelligence in shop term
balderdash and he echoes my position
about that we mustn’t cede our moral
authority to narrow AI in the short term
so ladies and gentlemen boys and girls
there is an opportunity here to make our
future to get the Utopia but it’s going
to need us to do things differently we
mustn’t blunder into the business of
usual prediction this particular graph
shows a 60 10% 30% for positive business
as usual so-called neutral and 30%
lacked ahead of a negative perspective
maybe maybe not
but the point is is that business as
usual is not good enough business as
usual as God as into the climate change
position it’s not good enough we must
become change masters we must build our
society we must take active control and
with that I hand over to the youngsters
in the audience because it’s you that
are going to make this difference into
the to this planet it’s you that need to
be engaged on this subject so excuse my
passion but I need you thank you very
much
[Applause]
Please follow and like us: