I want to thank him then and Europe
he had good news
[Applause]
grandchildren right there is a sign of
intelligence
I do not think it was just that
nicely programmed but keyword
intelligence what is the smartest
we are on earth when we are
intelligence is defined as the ability
to solve problems
then I think I could already here
a bit pondering
our cities are overcrowded
heat up the climate we lead wars
we waste the landscape
I think we need help and that
Good news is it is indeed
help on the way help from machines the
more and more approximate what
people can
Smart machines could be smart
Machines smart machines can
can walk can speak speak
can listen and see things
can read and write
can analyze sensor data and they
can learn by yourself
watch by imitating
rewards
I think that these smart ones
machines in their fields
can actually help
just play for the older
expectant population I think that
smart machines could be helpful
when cleaning up at home
maybe go shopping food cook the
Dishwasher give that
pour beer glass or even just
transmit messages or in the
industry smart industry known as
Industry 4.0 a high with us
automated very flexible production
Electrically networked autonomously at cars
These are the big trends of the cars
the advantage
there will be less traffic jams less
accidents and much more comfort too
Here again the old people
By 2050, the number of over
100 year old in Germany
tenfold
even 90 or 100 years old would like to
still mobile but they do not drive then
More auto auto you then call yourself
maybe an autonomous vehicle that
folds the seats out of the rising
comfortable 1 folds in the seats again
and they say the car should go
smart grids also one of these
tags today are necessary to
offer and demand in line too
bring in particular in the nets where
a lot of renewable energy
are present and these smart
Machines are already so good today
In some cases they are better than
People
I bring a few examples that is
the watson of ibm so in the middle
you see the earth with one
garland that symbolizes the
computer here at a quiz game
Verdi did a bit
more complex variant of who will
millionaire of this root system had in
his entire memory
internet lexicon wikipedia
he could read everything
he could get the information
to bring together and the special he
Made it into 230 customers and got it
he is actually the human
champions defeated but ibm wants
of course no quiz games
organize but use the
watson, for example, is referring to doctors today
help
because if the doctor has a cancer patient
then he can ask watson for advice
The Watson seeks thousands of people
anonymized patient records and
medical literature and tries that
Doctor helping with diagnosis and therapy
or also bank consultant stock transactions
For example, watson can help as well
in workshops at kreutzer
and so on google will mean it
maybe also known alpha that was
the software of those who have the
worlds best player the boardgame go the
World’s best player in spring this year
year defeated go is very much so
interesting because it is extremely complex
there is more on a board
Possible combinations as atoms in the
universe
that can no longer calculate computer
So alf ag had to learn millions
Examples of how to close well and
how successful they are next
example, kita times a mobile
robot runs faster than his
boot the human world champion over
100 and 200 meters distance
traffic sign recognition also such a
example
the best software for
Traffic Sign Recognition today has one
error rate of 0 5 percent so 99 5
percent are right and human
error rate is more than twice as high
examples from the industry wind turbines
Hundreds of sensors measure them
Vibration whether the turbine is running smoothly
and so on trains also have sensors
many many other machines though
you analyze what these sensors data
deliver then you can predict when
this system will fail and that
days or even weeks in advance if
The vibrations are not quite right though
it goes a bit wrong too
goes with the smart machines cancer
Recognizable learning software finds
Show more of cancer in
tissue samples so there is analyzed whether
a cell is a cancer cell or
no more indicative of cancer than in the
human literature at all
Of course, the doctors knew about it
google street view also help
an example
here are france driven in
france there are millions of
houses have the whole
house numbers
so take pictures of the pictures
houses the task was to read the
house numbers and locate the houses
the just when it’s people that
that would make about two
take years
The software has two hours
made
and another human example
emotion recognition typical emotions of the
people so base emotions as anger
joy surprise mourning show up
in the face and worldwide the same
regardless of culture regardless of
old and then it can of course
analyze a computer and the best
Systems are now even better than
Many people like you tests
has found that were only times now
seven examples of what marty machines
already better today than humans
and the development remains natural
do not stand this train does not stay
stand when comparing 1995 was the
the most powerful supercomputer of the
world could reach 100 billion
computing operations per second
that creates a good smartphone today
Today we carry in our bags the
Supercomputer from 20 years ago
And this development will happen again
We will continue in 20 years
supercomputers of today in bags
carry
and some people even say we will
in 2040 the processing speed
the storage capacity of a
human brain for under 1000
can buy dollars
then , so to speak, have second brain in
the bag who would not like to get off and
sometimes a second brain
it can go even further new
chips not by software
simulate how our brain works
but by hardware electrically the
Learn 10,000 times faster than ours
own brain, too, that is already
possible, however, still in the labs
and what comes to
These data are based on those smart
learning machines usually come
from the internet learning by the millions
of images of videos of audio files
They learn the cars look like one
piece of mozart sounds and so on
we have a data explosion in the
internet this shows that data
mountains of humanity so to speak and
smart man has calculated times
that until the year 2000, humanity
about two ex works on information
it has produced two billion
gigabyte
these two extra at the stand today
a single day
that means everything we do until the
year 2000 on data
we produce today in one day
smartphones with sensors of all kinds
one day and accordingly a lot
have opportunities to learn and
they can even learn without them
Data on the internet you can too
learn by watching people
by imitating people
trial and error through rewards
So my favorite robot is here
that it’s the heck off he goes for example
in genoa in the school actually one
small classroom for yourself and there
He learns from the human teachers
Learn how his toys look like
He learns how to clear the table
learning through play piano learned
to deal with part and bow
He is always learning very curiously
I have the osaka in japan again
met and he learns through
rewards
He will be rewarded for doing that
so that they help people
Can you look at everything then
Of course, one could also be afraid
get clear when these machines like that
then they could be smart
someday be smarter than us
killer robot super intelligence there is
yes, some scientists before that
to warn
I think it’s essentially science
fiction and it is also one more
Western science fiction we have in
West has grown up with
frankenstein with terminal with it
transformers
metal 9000 like all of them are called
East is very different in Japan
robots and smart machines helper of
People have always been in the
culture in literature it is one
Western history and when I go up
the researchers are asking the question
roboterforschung then say the all of
development of general intelligence
you would need a machine to us
it is still dangerous to be dangerous
very very far away so i think so
is something to which we at the moment
but at least have no fear but
here is the actually happen
this is about the jobs here it is
That’s why there are many studies which jobs
which activities in the future
be automated
and here is for example applied
different professions and the
probability automation in
Danger are actually routine jobs in
factories but also in offices especially
in offices this time bank consultant
insurance representative logistician
accountant easy cleaning staff bus
taxi truck driver when all
brings together about twelve
percent of jobs in germany where a lot
can be automated that would be 5
million jobs
but you have to keep in mind that are
activities that are automated
not necessarily the job that goes away
So a doctor has as explained by
with security rate machine in future
to help him with the
diagnosis but there are still the
human doctors and I think so too
not the trucks on the highways
will drive without that there still man
maybe they are doing it in there
something else meanwhile billing
but bring in new orders
will not be there in danger
Of course the jobs are very complex
tasks have been creative research
development of media design craft
which require high social skills
teachers social workers caregivers and
you always have to ask her is that
automation actually
economical and it becomes social
which does not always accept anything
the case is and accordingly
also decide to emerge
Of course a lot of new jobs so the
Example design of such machines
it’s like software development
35 years ago at the beginning of the 80s
There are no software developers so good
like none today since 19
millions of new jobs worldwide
arise
and I think in the end it will be like
said all life both people
but we will affect these things
really need
the smartphone was or start it
the smartcast come the smart grid
smart home smart transportation
smarthouse and was everything you needed
can imagine
we will need it for the older ones
expectant population we will
Need for the sustainable
energy systems for the
Mobility systems will need it
for liveable cities and for the
competitiveness of our industry
and I think we will not only
get used to it
we are actually in one
community between people and
smart machines live like we do today
quite naturally smartphones
take the course quite naturally
adopt social networks
that’s how we become in our lives
welcome
maybe not like the Japanese here
which is actually electronic or
Digital clones build out of robots
it does not have to be this, but we will be the one
encounter smart machines everywhere
that’s enough of the autonomous vehicle like
said right up to the chat at the
speaker boxes on our tables
which should come soon
we will see her everywhere and me
think if it’s right and if
we are not against the machines, so to speak
compete but learn with them
be sure to use them
reliable to use
also to ensure the safety
that they do not hurt anyone
if only to be so to speak
dance with the machines then I think
we have a good future there and then
we have the chance that the smart
machine together with us maybe
enable something that is smart humans
people could call that peacefully
the eco-conscious lives are the most creative
and that would really be the
good news thank you
[Applause]
”
Be First to Comment