Press "Enter" to skip to content

The Terminator, SkyNet and Alexa: The Present and Future of A.I. | Marc Talluto | TEDxIWU


simple question for everybody how many
of you feel like you interact with
artificial intelligence every single day
all right about five how many of you
feel like you are pumping information
into the cloud every day that could be
used to manipulate your life in some way
Oh fascinating interesting okay so
there’s a combination there so two
things I want to talk about or actually
kick this off the fourth Industrial
Revolution so this actually there’s a
quote from the current Google CEO sundar
Pichai that we are in an AI first world
the fourth Industrial Revolution I kind
of see it as the sneaky breakthrough
moment it’s happening it has happened
all the interactions as you’ll see are
actually heavily influenced by
artificial intelligence so let’s take a
look first there’s an old phrase we used
to use about having myriad data and a
dearth of information like I just I just
have a whole bunch of data but I don’t
really know what it means but let’s take
a deeper look data is the fuel of
machine learning how computers learn
about what they’re doing what they need
to do next how they should act how they
can win games let’s take a deeper look
first off make a separation between data
information and knowledge so there could
be an abundance of data but you really
don’t have any information if I was to
say the stock price of IBM today is
whatever 100 could you act on it you
would say well I don’t have enough
information I just have data points so
okay so data over time is going to give
you information on what you should do
but the knowledge is gained from
information over an abundance of time in
certain situations now you’re actually
thinking about other possibilities
outside of just the stock market think
about the data we actually produce so
there’s two different things to think
about one is the data of things weather
data all the time temperatures humidity
a little dissolve xn slater sensors
there are sensors on everything anybody
here have a Fitbit okay has anybody
heard the news recently of fitbit’s
being used to report on army locations
and movements around the world okay
sensors like Fitbit they’re everywhere
we are sensor izing everything smart
readers smart meters the Internet of
Things are all reporting data about our
lives smart fridges smart appliances our
utilities all of these devices even
engines on machines our cars they have
system logs they can all produce data
and you’ve ever been shopping on
something on Amazon and then you went to
your Gmail and there’s an ad for it
right we’re not quite creeped out about
it anymore right if this was 15 years
ago whoa how do you know what I’m
shopping for now I go yeah I’m going to
buy that but all of these things are
talking in the background let’s talk
about you two so now your age figure out
your gender might know your weight your
relationship status you give that up
freely your income you give to the
federal government I know by the census
data people know and can find out and
just say no the federal government gives
up income tax data not enough but
anonymously to data engines to figure
out what zip code what marital status
how much money you make so guess what
people can target market you and will
keep going every time you like something
every picture you take and share every
time you click through something every
movie you watch or pause or rewind
everything you purchase which credit
card you use or if you use PayPal is
being recorded and tracked so people can
figure out what to market to you what to
sell to you to make other influential
decisions and we willingly give all this
information up what’s really important
is this aggregation over time when you
put it in context you have results did I
win did I lose this is a fascinating
fact did you know that one there’s a
higher amount of smog the stock market
goes down it’s true it’s been proven
small correlations and how much fine
particulates are in the air around Wall
Street’s affects behavior and so on
those days stock traders aren’t as
risk-taking so the stock market declines
a little bit just a little bit but it is
noticeable and there is correlation so
this creates information information
that is actually extremely useful I have
to make this comment because our
cybernetic overlords are coming and they
might already be controlling us but well
we’ll get into that
but there’s massive acceleration in not
just the computing power but the ability
for the computing power we have to
discern through all of these piles of
information we’ve been producing in
order to make meaningful decisions for
us big data if you haven’t heard of this
term it’s kind of what I was mentioning
before we have stockpiles of data but we
don’t necessarily know how they’re
correlated but we’re saving it all we’re
storing it all does anybody know the NSA
is storing every communication every
cell phone conversation every email you
have in a two billion dollar facility in
Utah and I swear I don’t believe in
aliens I’m just telling you it’s
actually happening in the world they are
they’re storing all this information so
we have these pools and data data that
we sometimes we don’t know if it’s
relevant but we know we have it feeding
this is the Internet of Things if you’ve
heard of this every device smart
televisions your phones of course we are
all basically giving up our location
data or giving up what we wear we’re
driving how fast we’re going you can
argue that your device is an Internet of
Things for people but the Internet of
Things also applies to like I mentioned
smart meters cars we just have
essentially all these sensors reporting
on what’s happening in the world now
robotics is an applied application to
using this information in more the
physical world artificial intelligence
is what I’m here to talk about this is
taking all of that data allowing
computers to actually program themselves
to learn things and we don’t necessarily
artificial intelligence the difference
between artificial intelligence and
machine learning is that we don’t know
necessarily what the results will be we
program the computer look for things to
make correlations so try to figure out
something we don’t actually know we
don’t point it to go play checkers we
pointed an information set and say and
say tell me why stock market dip so I
should even say stock market dips and
they say you know we found out there
were smog on that day things like that
that’s where artificial intelligence is
going and it’s even going farther so
what’s good let’s take a look a moment
and think about it every does everybody
not use Netflix or Hulu or something
else okay let’s think about how you used
to go to the movies you used to buy a
generic ticket and the theater didn’t
know who you were that went about what
the movie theater is what the studios
knew was
that somebody bought a ticket to see a
particular movie somewhere in the world
at maybe at that particular movie
theater that was kind of it and we’ve
seen movies bomb we’ve seen you know
there’s been bad movies what Netflix
knows about you okay sorry who you are I
know what you watch I know when you
watch it I know when you start a series
I know if you finish a series I know
what you like them didn’t like because
we put three stars five stars we
actually put thumbs up thumbs down I
know when you now Netflix knows there
are five people in your family I can
tell when you log in and watch a series
but you’re your companion you chose to
watch it without them at some point I
can see that this is actually happening
so what does Netflix done with this
information
has anybody seen house of cards stranger
things they’ve gotten into original
content they don’t make things that bomb
why because they already know what you
like they already know what you’re
interested in and they know who to
market it to I know your age I know your
location I know how many people are in
your family I know that my kids watch
way too much on Saturday morning that is
the difference that’s how this
information starting to be used so this
is still the good so let’s move forward
so automated farming vertical farming I
know how much water I’ve given a plant
cameras on these plants I know when
they’re actually growing I know what the
proper temperature is I’m using all
these sensors kyoto japan currently has
a plant that’s producing ten million ten
million heads of lettuce in a given year
through vertical farming using
artificial intelligence driverless
vehicles some people that argue dry
texting means it is a driverless vehicle
trust me I do it myself but actually it
is this is real it is happening okay so
the car manufacturers are doing a GM
Ford Google Apple uber lyft cities
across America these cars are driving
presently Tesla and not only is Tesla
doing automated driving they’re sharing
the results of driving and where to turn
and what they did that one car what it
does
Tesla actually reports back to
the mothership and that makes
improvements to its software so all the
other cars suddenly learn from what you
drove and how you drove
so the actual auto driving mechanism and
Tesla is self learning and it’s adaptive
day by day ah the medical field
nothing is more awash in data about you
your weight your blood pressure you know
certain diseases may be genetic diseases
other things you may have but like for
MRIs scanning MRIs scanning one if you
wanted one doctor to scan one our Mir
MRI what would you like to see a million
MRIs scanned and try to find the pattern
between them so they can determine if
yours is at risk for a disease currently
computers can has a 90% accuracy on
diagnosing skin cancer there is no
doctor that is that good this is the one
that is pretty impressive sorry alphago
alphago if you’re familiar with chess
and checkers like I mentioned go is a
game a little bit more sophisticated
more strategic so Google said they said
that could never be beaten by a computer
human couldn’t be beaten by a computer
not only did alphago beat a human
actually two out of three within life he
was in the last two years they then
built alphago zero which actually did
more deep learning more artificial
intelligence to train itself it then
beat alphago a hundred times straight
then they pointed alphago zero at a
chess basically just how to play chess
they just gave it instructions within
eight hours it beats it beat the world’s
best chess computer so that sounds
positive that seems fun right the other
side of this is the flash crash now this
was eight years ago
high-frequency trading algorithms if
you’re not familiar with this stock the
stock market is not controlled by humans
saying I want to buy stock they are
controlled by algorithms that are
programmed to use artificial
intelligence to say mmm it’s there’s
weather here this is gonna affect butter
futures I see information about the
Nikkei index how is that going to affect
the stock price the problem is those
bots actually trade against each other I
actually came out of the high-frequency
trading business years ago they actually
put simulated fake trades to see what
your bot is gonna do against my bot now
my trains are battling it
this flash crash was actually the Dow
lost 10% of market value for about 12
minutes back in 2010 they say 60% of all
trades on stock markets are actually
done by high-frequency trading
algorithms I wouldn’t want to be killed
by a robot but in the Geneva where they
have the Geneva Convention just last
month
autonomous lethal robotic systems is a
big concern we’re familiar with drones
but the actual button that says I should
drop this bomb it still has human
intervention they’re now trying to
prevent those going to the mind of a
computer say well I see an intruder I
should shoot this person it’s kind of
like Robocop but this is this is very
real this is both with air warfare
missiles and ground-based systems there
is if you’re familiar with Elon Musk who
runs Tesla Bill Gates with Microsoft
Stephen Hawking the famous physicist
they’ve all been warning extensively
against a robotic future of warfare the
strange thing is you know who leads in
the development of autonomous weapon
systems the United States do you know
who then is following suit because we’ve
led the charge China Russia
high-confidence Russia interfered
whether or not there’s collusion is not
the question here the ability for a
foreign nation to buy ads to make sure
what you click on and what you like
influences what you see and how you
perceive to then make you act and do
something that maybe you wouldn’t have
done is an effect of artificial
intelligence so if you’ve seen Facebook
in the news is changing their newsfeed
they’re trying to they’re trying to dial
back actually all that automated news
that you’re receiving I don’t know this
is kind of leads me into the weird there
is an AI religion I kid you not it’s
funny it’s actually WTF but I don’t know
what to do about that
but the way of the future basically
saying that intelligence shouldn’t be
just limited to a biological
intelligence will we not have a higher
power that has all this information all
this calculation power isn’t that a
higher intelligence that we then almost
become robots for so now the weird
I can’t explain some of this stuff but
the best restaurant in London by
TripAdvisor
didn’t even exist that’s right some guy
thought it would just be funny to create
an address create a phone number put an
ad online and he game the system people
would call him to make reservations he’d
say now sorry we’re full sorry we’re
full I believe people gave good reviews
to it and kept pumping it up pumping it
up it was the number one reviewed
restaurant in London it was his literal
shed and it did not exist so you may
heard the story of Alexa dialing 9-1-1
so there’s a domestic domestic violence
abuse and somehow the cops just showed
up and they said well Alexa wasn’t able
to make a voice call like well somebody
called 9-1-1 and we can actually hear
the arguments so somebody in this home
dialed 9-1-1 so it’s like that Alexa
just intervene no tell you the Microsoft
AI that they put out there to
communicate as a chatbox became a racist
and a sexist within eight hours and they
had to shut it off language creation so
they had AI BOTS actually talk to each
other and chat to each other they
started talking in their own language
they were still using English but then
they started to use their own
manipulation of English to actually make
it faster to talk to each other Elon
Musk as I mentioned before is actually a
big he’s very much against the the
future of AI because he’s very much
concerned what the impact could be the
humanity so he’s actually built a
company called neural Inc they’re just
forming to help build brain computer
interfaces for humans so he can take all
this interface and we could take all
this information and actually use that
information ourselves how I’m not too
sure there is the IAI creative arts
there is music here’s where it gets
weird if you get run over by a
autonomous car if you get shot by a
autonomous gun who’s really to blame is
it the company that built them is it the
programmer like you know the entire
legal system is based on somebody being
accountable for that action now it’s
actually artificial intelligence one
thing about the fourth Industrial
Revolution is that AI has been said to
increase GDP by 17 trillion but the work
forces of advanced economies they rely
on those financial job
like I mentioned that are going to be
easily actually easier it easily
replaced and this is a 2030 estimate so
we’re talking only in 12 years you have
high skill high pay jobs actually that
will still be promoted people that
actually be coding these things maybes
more sophisticated type of work that
can’t be replaced by AI but really low
skill low paid jobs this disparity will
continue to increase we always talk
about the wealth gap and people who
recovered from the last recession this
is just going to escalate even more and
more the upside is there will be we talk
about a massive elimination of waste in
the supply chain to know who needs what
and when and where the logistics behind
this is becoming far more sophisticated
I’ll have to say this too this is not
something we’re going to stop it isn’t
the benefit of every company to do this
to eliminate this waste
there is really no stopping the progress
that we’re making in artificial
intelligence
just the thought think about what we do
to try to prevent nuclear weapons from
giving out of hand if you had a
proliferation of artificial intelligence
based weapons that are actually smaller
and easier to smuggle but some despot in
some country could easily you know
suppress people with it probably a
higher proliferation of those weapons
one thing I will say is China is betting
bigger than we are
so the impact that on the US China will
actually have a double their GDP impact
because they’re making a bigger bet on
artificial intelligence so let’s think
about this for a second
AI tools are available for everybody to
use Google has and Facebook azzam you
can go out there open-source you can go
on the cloud right now and start using
them these aren’t mystery programs that
people you know hiding somewhere and why
because they want your data the more you
use it the more they’ll collect your
data so if my choices are already being
selected and filtered the news I choose
to read the routes I choose to drive the
movies that are being chosen for me if
I’m being influenced I often get fake
data potentially like that restaurant if
my decisions are being based on what I
read and experience my education my news
can be faked
right but it is being selective for me i
artificial intelligence sometimes even
written by healy effects the food it’s
actually feeding me those farms the
Logistics I’m kept alive with that early
detection to make sure I live nice and
long but at some point I could be given
false data my Apple map or actually
don’t use Apple Maps my Google map could
drive me off into a bad part of town
I could be shot by an AI robot am I
already living in an artificial
intelligence world that’s all I got so
thank you
Please follow and like us: