IEEE Robotics & Automation Magazine - March 2011 - 37

about the give and take of human relationships. Similarly, a
robot is not going to be an adequate replacement for a parent
in terms of an infant's linguistic development. Advances in
natural language processing could lead to superficially convincing conversations between robots and children in the
near future. However, such interactions would not be meaningful in the way that caring adult-child interactions are. It is
one thing for a machine to give a convincing conversational
response to a remark or question and a completely different thing to provide appropriate guidance or well-founded
answers to puzzling cultural questions. There are many
cues that an adult human uses to understand what answer
the child requires and at what level.
Language interactions between young children and adults
are transactional in nature, both participants change over
time. Adults change register according to the child's abilities
and understanding. They continuously assess the child's
comprehension abilities through both language and nonverbal cues and push along the child's understanding. This
is required for both language development and cognitive
development in general. It would be extremely difficult to
find specifiable rules that a robot could apply for transactional communication to adequately replace a care giver's
intuitions about appropriate guidance.
Babies and infants would probably not be able to resist
interacting with a robot that responded to them contingently. There are, however, reasons to fear the effects of
such interactions, given that an infant's experiences of
interactions have such a powerful effect on their development. What an infant learns about getting a response from
a robot nanny is unlikely to help it understand the subtle
and nuanced reciprocal interactions that are needed to
form good human relationships. It might seem convenient
to have a robot entertain your baby so that you can get
some more work done, but the risks might be too great.
In addition to impeded social, emotional, and linguistic
development, a young child spending too much time with
a robot might suffer other negative consequences. Bryson
[33] points out that interactions with robots will be much
more predictable than interactions with humans and that
children might come to prefer this. In a related argument,
Kubinyi et al. [34] argue that just as cross-fostered animals
and birds learn behaviors and responses when raised by
those of a different species, so humans raised by robots
might develop differently. They might, for instance, grow
up dependent on individualized entertaining systems and
be socialized to follow nonhuman behavioral patterns. A
new form of human, homo technicus might emerge [34].
Melson [35] also considers the effects of adapting to pseudointeractions with technology and suggests that if children
begin to think about robots as being alive, they may also
begin to think about humans and animals in more mechanistic terms and with less regard to their moral standing.
There are considerable risks of negative consequences
from leaving babies and infants in the company of robots.
The same is not necessarily true for older children. For

children who have formed secure attachments to human
caregivers and who have a good grounding in human-social
interaction, some exposure to robots might even be useful.
Since robots will probably play an increasingly important
role in society, it would be just as well if children were educated about their workings and familiar with them. Melson
[35] suggests that robotic literacy should be encouraged for
both parents and children. "Such 'literacy' would help
adopters of this technology understand: 1) how robots are
produced, maintained, and operated, emphasizing their
human-produced properties; 2) what the limits and potentials are for various robotic technologies; and 3) what the
distinctions are between living and 'pretend' living-stuffed
animals, puppets, and robots." In addition, encouraging
children and adults to understand the nature of anthropomorphism and the methods that can be used to strengthen
the illusion of mental states in nonliving machines could be
a powerful way of protecting them from the ill effects that
might result from overestimating the abilities of robots.
Conclusions
Clearly, there is a growing interest in developing robot
caregivers and companions, particularly for the youngest
and eldest members of society. At the same time, there is an
ever-increasing ability to implement design features that
create the illusion that robots are sentient and able to respond
emotionally to us. Such developments raise the likelihood
that vulnerable members of society will be left in the company
of robots and that people will mistakenly believe that the
robots are capable of caring for them and forming mutual
relationships. In this article, we have probed the ethics of
designing robots that promote the illusion of being able to
form meaningful relationships with humans.
It is acknowledged that some form of deception is involved
in developing robots that appear to understand us. However,
this deception depends on exploiting the natural anthropomorphism of the observer. Anthropomorphic design is prevalent in many aspects of society outside of robotics, and to an
extent, being anthropomorphic may be an unavoidable part of
being human. Clearly, it would be unreasonable to call all such
design unethical. Our arguments are focused on cases where
the probable consequences are a reduction in well-being.
It is suggested that, for various reasons, the young and
the elderly are likely to be particularly susceptible to such
designs. We conclude that robot companions for the elderly
could offer positive benefits in terms of improvements in
health and welfare, although these are risks in terms of dignity and loss of social contact. In contrast, the development
of robot companions and caregivers for babies and infants
are more likely to lead to negative consequences. The attachments that infants form with human caregivers fashion the
basis of their emotional and social development, and infants
that spent too long interacting with robots could learn aberrant forms of interaction. There are reasons to be ethically
concerned about the possible effects of exposing either of these
vulnerable groups of society to robot care and companionship,
MARCH 2011

IEEE ROBOTICS & AUTOMATION MAGAZINE

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - March 2011