IEEE Technology and Society Magazine - March 2020 - 57

between the robot and the human [16]. In human-centered computing we recognize the primacy of the
human and the purpose of robot in as grounded in
human flourishing.
But if we take a post-human perspective, we
demote the human to being no more important than a
rat or a robot and we will be liable to act unethically
because we see no difference between the machine
and the human. It seems to me that social robotics
focuses too much on the robot and not enough on the
human. The robot should be viewed as a support
worker for the human.
Technocentric approaches are apparent in robot
therapy. In the absence of the robot, the focus is on the
child: psychologist, teacher, and parent sitting looking at
and engaging with the child. The conversation involves
the child. When a robot is introduced, the space changes. The robot is central, perhaps raised on a table. The
child, parent, and therapist sit facing the robot. The conversation is focused on the robot and its behavior. While
this may have certain advantages in freeing the child
from intense observation and thus decreasing anxiety, it
is also dehumanizing, reducing the child to an object,
manipulated by a superior robot. If the interaction subjugates the human and makes him subject to the commands of a robot, this is a matter of ethical concern.
Of course, if we value the human we will protect his
identity, promote as much autonomy as possible, and
conduct the relationship in complete honesty. The value
we put on humans will depend on how we view their
relationship with technology. The technology should be
a help and support within a social community.
If we look towards the singularity [11], when artificial
intelligence becomes superior to humans and supersedes humans, we will be likely to raise the robot above
the human. A post-human perspective will raise the status of the robot to equivalent to the human and then
superior. We will deny the personhood of the individual
who, as Lovass considered, becomes a biological husk
[4]. And in relationship terms, in deploying a robot, we
are declaring ourselves as superior to the human, creating an unbalanced relationship, and taking on godlike
capacity to program the human. The relationship then
expressed with the autistic child is then one of domination not dialogue.

Empathy
An adequate social robot would interact with the person, not the social label placed on them, or the external
perception of their behavior. So the requirement is to
understand the child, who they are, how they feel and
think, and not just treat the child generically as autistic.
This level of interaction requires empathy and
human understanding rather than programmed
MARCH 2020

∕

Use of robots should be evaluated
based on purpose and process,
not on the technology itself.

responses. It is well beyond the most sophisticated
social robot. In addition to the complexity of the interaction required, the requisite variety of response
needed is beyond programming. The interaction cannot be rule based. It is interactive, interpretive, tacit,
and difficult to pin down.
Empathy towards autism involves understanding
the child's outlook, character, skills, and unique difficulties. The autistic child will process information differently to a neurotypical person. We can easily ignore
their thoughts, feelings, and perspective and radically
misinterpret their intentions. We can interpret withdrawal as resistance to us, as spiteful behavior, rather
than an attempt to cope with sensory overload. Processing communication in daily activity can be very
draining. Some interactions that might be easy for us
can be hard work for an autistic child. And we can end
up being less than empathic towards the child.
So empathy and understanding the viewpoint of the
child is not only a virtue, but essential for any ethical
framework. We not only have to empathize with the
child, but we need to consider the impact of the robot
on the child. The view that the robot, as a simple
mechanical object, is going to be preferred by the child
may not be true.
It will depend on the context: the carer or parent
may be present, the researcher may intervene, the environment may be different and challenging, the robot a
focus of anxiety. We may think the child is enjoying the
robot interaction when actually the child is aware of the
expectations around him and trying to fulfill them.
We may have no understanding of why an autistic
child suddenly reacts aggressively to a robot, or what
actually triggers successful robot interaction because
we have made little effort to empathize with the child.
The autistic child sees, feels, and thinks differently to
the neurotypical person. Their sensory overload may
not be eased by the added burden of interacting with
a robot in a complex environment.
Rather than pursuing the ethics of the robot we
should consider the ethics of the autistic child. Rather
than spending large amounts of time studying the
robot, the robot researcher should spend large
amounts of time studying autism and being with the
autistic child. Having understood the autistic child and

IEEE TECHNOLOGY AND SOCIETY MAGAZINE

IEEE Technology and Society Magazine - March 2020

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - March 2020