[Applause]
so this was an official video from the
2015 DARPA Robotics Challenge now
objectively speaking we saw a number of
humanoid robots who tried to do normal
human tasks like walking opening a door
or getting out of a car but they failed
now people usually have a different kind
of experience they are empathizing with
these robots they might even feel sorry
for them now I’ve been working with
robots now for more than 20 years and my
background is in biology and Natural
Sciences but as soon as I moved into
robotics I realized that there’s a
fundamental differences in the nature of
robots and people or biological
organisms so this is just a picture of a
robot that I did quite early in my
career these days I’m working with quite
complex machines back at University of
Hertfordshire no robot house we develop
technology that can help people older
people to live independently for longer
in their own homes so people who can
normally still do most things but might
have some problems with physical tasks
or with cognitive tasks so this is one
of the robots we work with this is
another robot so these robots are social
people interact with them and the robots
can for example fetch object for them or
remind them to take medicine or can
remind them of appointments now for me
robots are machines if they break we fix
them or we replace them and when our
work is done we switch them off I don’t
feel guilty about that but most people
or yeah most people have a different
attitude towards robots they very easily
empathize with them or treat them as if
they were creatures little animals or
even people and that’s the topic of my
talk I’m where I’m questioning a little
bit why is that so why do we have this
this reaction
so in many debates robots have been
compared with people for example just
recently literally I think a couple of
weeks ago European members of parliament
voted to give robots legal status as
electronic persons with rights and
regulations now it is very important to
have discussions on legal ethical and
social such social issues but these
discussions usually only point out one
trajectory of the future of Robotics
namely to make these robots more and
more humanoid humanoid and more and more
human-like in their behavior and their
appearance and their intelligence for
example what you see on the right here
is an Android built by Yoshi Ishikawa in
Japan in his own own image so let me
first explore a little bit why are
people interested in viewing robots as
as creatures why do they emphasize with
them first of all we have a deep-rooted
desire to imitate nature starting with
cave paintings this depicting animals
depicting people later in ancient Greek
and ancient Egypt for example people
were very interested in building statues
that could seemingly move as if they
were alive using basic technologies that
they had at the time namely ropes and
water and fire in the 18th and 19th
century for example here in Europe you
had inventors building so-called
mechanical automata basically based on
clockwork technology that was available
at that time so these were animal-like
automata that could replicate these
movements some of the movements of the
animals or they looked human-like like
what do you see in the in the middle of
the slide and these human-like automata
they replicated human behaviors such as
writing dancing or playing musical
instruments now these days we have a
different level of technology we have a
lot of computer and robotics technology
but something similar is happening we
are interested in replicating nature in
order building robots in our
own image and make them do things make
them do tasks that are human-like
although you saw in the initial video
that they may fail at doing that so
there’s also the way in how we interpret
seemingly intentional movements so this
is a video from a study from 1944 done
by two psychologists called Heider and
Simmel what they did is they showed this
video to a group of people and with no
other instructions and just afterwards
asked them please write down what you
saw that was it
now what you do see objectively speaking
are geometrical shapes two triangles a
big one and a small one and a circle and
they move around and in and out of
another geometrical shape but this is
not what most people said they talked
about agents they talked about
intentions they talked about oh there’s
little triangles right just sneaking up
on the other ones they they talked about
these geometrical shapes chasing each
each other having goals hiding from each
each other so we just can’t help
interpreting it seemingly intentional
movements in this way in a human-like
way in a creature like way it is also
Jerome Bruner another psychologists has
said and written many many times that
how we interpret and how we make sense
of the world is in terms of narratives
narratives with intentional agents and
that something we also apply to to
robots now why was it important for our
ancestors to have this acute sense of
recognizing an intentional movement well
it was very important to distinguish
intentional movement form movements
caused by physical processes for example
clouds moving in the sky which don’t
pause which don’t pose any immediate
threat but it’s important to recognize
creatures because creatures can be
either predators or prey or mates so we
are tuned we biologically tuned in
recognizing intentional movements and
that the same applies to robots so
here’s a is a little mobile robot
and you know the way how we interpret
the way how how we view this is that we
automatically wonder what is this robot
up to what is it actually doing you know
it must be doing something but this type
of behavior when our reactions to the
robots it doesn’t make these robots more
human-like it doesn’t make them
creatures there’s a fundamental
difference between living things and
robots robots are machines they are made
by us any robot is more similar to my
smartphone or my toaster than it is to
you or me now another aspect that
influences how we view robots is their
design this is the power hole robot
design partic on ovc beta in Japan it
has been used for many years
successfully in homes for elderly people
and people with dementia its design is
based on a baby seal robot using this
baby schema with a big head and big big
eyes the same as you stand you know
cartoon figures to make things look cute
real baby seals are cued and so so it’s
this baby robot that’s how it is being
perceived it has been designed with that
purpose there’s no other purpose to it
rather than interact with people to
motivate them to engage them to also
motivate them to interact with other
resonance in the home so this robot is
an example that you can without building
a humanoid robot you can find an
application where robots of other shapes
can very successfully be used in a
particular application area more
recently we find these very abstract
shapes these more robot like devices
that you can communicate with like
Amazon echo which is again an example
you can have something that is
absolutely not human-like but it can
provide useful services in a particular
area now we also have an example of a
humanoid robot called Casper that shows
the benefit if you introduce a robot
specifically as a mer
not pretending that it is human-like
Vespa has been designed to interact
specifically with children with autism
it doesn’t look pretty if that’s what
people tell me but the reason is it has
been designed for very particular
purpose it has the face for example has
realistic human-like features but very
much simplified to help children with
autism because they are often
overwhelmed by the complexity of human
interaction and communication with the
robot on the other hand they can in
their own time
explore interaction if they for example
hit the robot then the robot will react
in a way that will show the child what
the consequences are of his or her own
own actions and then a teacher or a
parent can use it as a as an example to
discuss what is socially appropriate
touch why is it bad to hit somebody you
could not do that with her with a human
being and not even with a robot where
people think it is a human but you can
do it with a robot that specifically
introduced as a as a machine so we treat
robots as creatures and humanoid robots
very often as if they did they were
human and there was nothing wrong with
that this is how we are this is how how
I our body works how our brain works how
we perceive and interpret the world
however we must be aware of the
fundamental differences between
biological and artificial systems humans
have an evolutionary and developmental
history we are born into a society where
we grow up as social beings and we learn
about the world and ourselves through
interactions with other people what was
have nothing like that they don’t have
bones they don’t have blood they don’t
have any genuine emotions so we need to
be very clear even if we treat robot
socially if we treat them like creatures
or in a human-like way we need to be
aware that this treatment doesn’t change
the nature of what they are they are
so the question is now do we limit our
imagination to this one possible to
reactivity of the future of robotics
that goes into ever more human-like
systems or do we under one on the other
hand broaden our imagination and say
well there may be many many possible
futures let’s just imagine different
roles that robots can play usefully in
our society where they fulfil certain
roles and let’s just match these vaults
with specific robot designs in the way
in how they look how they behave and how
intelligent they are so these mappings
they should determine the future of
robotics robots are not people but it is
up to us to decide what they may become
thank you [Applause]