Press "Enter" to skip to content

Will care robots care? | Astrid Weiss | TEDxTUWien


[Music]
so let me start with the good news first
the good news is that from 2000 to 2015
global life expectancy increased by five
years this gives all of us more time to
check of things from our ten thousand
things to do before we die bucket lists
but what if what if we are not capable
of maintaining independent living at
home when we are old
what if daily activities such as
cleaning cooking managing medication
become too challenging for us or even
simpler tasks eating walking there is
this vision of robots living together
with us in future in our homes helping
us with these daily activities and then
we again have time for our bucket lists
what would you expect of such a robot
would you want to have a machine just
the machine or would you want to have a
companion that really cares for you I’m
working in a research field named human
robot interaction maybe not everyone is
familiar with my job is to study how
humans react towards robots I’m not
building the next Terminator I’m not
interested in artificial people or
modern working slaves I want to know how
people react to robots and I’m curious
how they must be designed in future to
serve as if reasonable purposes so the
emergence of so called social service
robots at the end of the 20th century
was a departure and the way how we
considered machines once of a sudden we
were interested in having machines
acting in place of people in specific
social roles such as servants or
Butler’s like the Cairo bot developed a
loofah in Germany or working
as a teacher teaching your kids foreign
languages at school like the NK robot in
South Korea or working as nurses and
hospitals lifting you in and out of your
hospital bed like the Reba robot in
Japan or even a sexual partners and that
pick two the last one is a stock picture
if you’re interested in real sex robots
google it yourself after the talk to
make this vision of social service
robots reality roboticists are placing
huge emphasis on the appearance of
machines so the idea is if a robot
should be social it must perform in a
social way using so-called social cues
social cues are hints that let us human
reply to something and in a social way
its hints that let us automatically
react toward something and because of
that many believe that robots need to
have faces arms bodies something that
resemble the human or provide a really
convincing simulacrum and just for
clarification reasons the sitting guys
in that picture are the robots
so I partly agree with this view so
clearly we react more likely social
toward something that has a face or what
we know what to speak to you however can
it be only just about looks this way of
designing things is also called
anthropomorphism but is it only just
about looks ten years ago when I did my
PhD in sociology I was totally convinced
– what I’ve learned in sociology that
social interaction is only but only
possible between humans no exclusions so
when I first had to tie the chance to
work with social robots during my PhD
and to check how people reacted I became
curious okay what does social actually
mean
to us or in robotics and in everyday
language we use the word in a variety of
different ways so let us have a quick
look on that we call it social where we
go to church church is a social
institution we consider everything that
happens in institutions or organizations
as social such as doing sports together
it’s a social thing we consider
something like high society is existing
that difference is in your degree of
social nets and magazines of social
gossip columns we even call ants
social insects because they live in
colonies and finally in everyday
language what we most likely consider a
social is friendly companionship the so
called social cup of coffee we enjoy
with someone and if all that means
social can robots be social and robots
become companions for us actually care
for us can they be social so in the
mid-90s two professors at Stanford
University called recentness
wrote a book the media equation and the
book is about a theory of the same name
this theory derived from a series of
social psychology experiments they
performed and these experiments show to
sum it up in a few words that we humans
tend to treat media technology computers
phones television sets similarly like we
treat other humans in other words the
rules that apply for us in social
interaction when we interact with other
people equally well apply when we
interact with technology let me give you
an example recent less developed
computer programs and they gave them
behavioral cues from extrovert and
introvert behavior and then they let
people use these programs in Austin
afterwards so what do you think about
that program and it turned out that
people prefer people with extrovert
behavior traits preferred extrovert
computer programs and introvert purple
keep up
preferred introvert ones and the
interesting fact is this is pretty
similar like how we treat other people
we prefer people that are similar to us
more than people who are different from
us and reason that’s could even show
whether their studies that even if we
think it’s stupid it doesn’t make any
sense it’s not reasonable to treat a
computer like another human we we simply
do it it automatically happens it’s kind
of in our blood it’s a one thing where
my sociology studies were right was that
social interaction is a deeply rooted
human trait however communication media
such as computers televisions they don’t
look too human-like so what about robots
is it really just about the looks of
robots that makes them social I give you
an example this is the Hobbit robot the
hobbit robot was developed in a recent
EU project I was involved in and it was
developed especially for a case I gave
you in the beginning namely an abling
all the adults to maintain independent
living at home so the robot could
navigate for apartments it could search
if if the user is there it had an
entertainment functionality it had
Fitness functionalities it could remind
the user of drinking enough water during
the day all this stuff of
functionalities but the crucial idea was
that the robot has something called
mutual care behavior interaction
paradigm sounds fancy means that it’s
not only the robot taking care of the
old adult but it could also work the
other way around that the older adult is
taking care of the robot which is
helpful on a technical side as well for
instance when the robot gets stuck and
the user helps or rescues the robot
ethic and continued navigation but the
main idea was when you have this mutual
care it creates kind of a bond over time
between the older adult and the robot
and that should raise technology
acceptance on
long run this kind of sounds too good to
be true right it’s it’s just a matter of
time it’s a matter of technical advances
and at one point we will have our
service robot at home helping us and we
can go back to our bucket lists but will
they really care and is it important if
they really care how will the human
robot relationship of the future look
like when we really have them at home
sharing everyday life with us and the
movie robot and Frank especially such a
scenario is targeted Frank son hunter no
longer wants to take care of his father
who’s dementia is increasing so he buys
him a care robot what else you see what
the Hobbit project was kind of intending
is pretty close to what Reese what
science fiction is imagining in future
so the ideas are not that far apart
however in the movie Frank is not
bonding with the robot because of a
mutual care interaction paradigm Frank
is bonding with the robot when it turns
out that the robot cannot distinguish
between legal and illegal and Frank
teaches the robot to help him with some
burglaries and that creates the bond
right but that’s also another storyline
in the movie
namely the storyline of Frank’s daughter
medicine and medicine is a firm believer
that human care is better than robot
care so she wants to take care of her
father instead of the robot problem is
she’s totally overwhelmed with the
situation she is that much overwhelmed
that at one point she decides to
secretly switch on the robot to help her
clean the apartment after the mess her
father created the robot is not just
servant you cannot switch him on and off
like he is a slave the robot is my
friend
so that’s Frank’s reaction towards
medicine secret activity and these
potential effects of future human
relationships are the most interest
sting ones from a sociological or
ethical perspective robots are not
people
robots are machines machines are not
sentient robots can express emotions
they can pretend to bond with you but
these are all simulations these are not
shared experiences like we humans have
emotions that a robot shows may look
human-like but they are lacking
situational understanding their
pre-programmed behavior but does that
matter doesn’t matter fella care
behavior is it important if the robot
really cares for us let me play a short
thought experiment together with you
which I developed with a dear friend
Bert Baumgartner University of Idaho a
philosopher we have Frank the elderly
man is the care receiver we have hunter
Frank’s son medicine his daughter and
the robot all potential caretakers of
Frank and the potential caretakers can
have some kind of internal state which
is the emotion intention motivation they
have to care for Frank and this day can
be positive or negative and then we have
Frank who can be satisfied or
unsatisfied with the performed care
behavior so this gives us a four field
matrix let’s have a look at the fields
with hunter Frank’s son who no longer
wants to take care of his father he
considers as the burden a duty he has to
fulfill so in the end he even buys him
the robot and what about Frank Frank is
not satisfied with this care behavior so
Frank is unhappy or what about medicine
medicine we saw she has a really
positive internal state she really wants
to care for her father but you cannot
handle it he’s overwhelmed so Frank is
unsatisfied with the care behavior and
result Frank is unhappy let’s have a
look at the robot the robot for instance
cooks healthy meals for Frank so we
considered a dapper
a positive internal state right the
robot really cares for Frank’s health
and Frank is super satisfied he no
longer has to take care of his eating
routine himself so Frank is happy and
then we have the robot helping with the
burglaries and human like we would say
well that’s that’s not a closet that’s a
negative trade you you don’t do
burglaries that’s not a good thing to do
right but Frank is super satisfied with
this care behavior and really happy
about it
what I want to show and tell where this
is the internal state doesn’t matter
what matters is the outcome of the care
behavior and if the care receiver is
happy with it that’s the final relevant
step is the performed behavior of the
robot the behavior that leads to a good
care behavior but clearly ethical
dilemmas remain is it deception if a
robot shows emotions expresses emotions
in my opinion it is it is if you are not
aware that the robot cannot feel real
empathy it is similar deceptive as if
another human pretends to really care
for you and in reality doesn’t but also
in that case it should be important to
you if the outcome the care behavior is
the right one that should be teeth and
not the internal state the thing is a
present human robot interaction research
cannot answer the questions how the
future relationships will look like
we’re lacking the data we don’t have
long-term studies over years and decades
and to achieve the data is really
difficult and risky and if you think
it’s about humans who really depend on a
robot it’s not easy to get that data but
even if we have it the answers will most
likely not be black-and-white there are
so many variations how humans human
relationships look like there will be
many variations how human robot
relationships look like however human
robot interaction research is necessary
if we want to develop social service
robots that serve us with reasonable
purposes in future but don’t get into
the trap that the robots are empathetic
there are pre-programmed machines they
should serve you and they use social
cues and social dynamics in order to
produce the right outcome for you and
the robot doesn’t have any intrinsic
motivation to be a friend thank you
you
Please follow and like us: