Transcript EECS 690

EECS 690
March 29
Robots as soldiers
• There are distinct advantages to the extensive use of
robots in combat, and advantages to more autonomous
robots. Specifically, a nation may preserve the lives of
its own soldiers by the use of robots in combat, and
greater autonomy means less reliance on large numbers
of specially trained personnel.
• However, some concerns suggest themselves as well:
– Determining the acceptable level of civilian/friendly casualties
– Increased use of military force because of less risk to one
nation’s soldiers.
– Re-evaluating the cultural assumptions of warfare.
– The possibility of making warfare more destructive.
Robots as sex objects
• The authors don’t say much about this, but
point it out as an area of technology likely
to be developed. They also point out that
whether such uses of robotics technology
have a tendency to lead to aberrant or
antisocial behavior is an empirical
question.
Robots as slaves
• A more general form of the previous two
examples is just plain robotic slavery. With
sophisticated robots, it appears that humanity
could get all of the benefit of slavery without the
moral cost. Again, some concerns suggest
themselves:
– Could this lead to persons who cannot afford the
technology reinstituting human slavery to compete
with those who can afford the technology (much as
sweat-shops compete with robotic assembly-lines)?
– Could this lead to human sloth?
Specialization
• Artificial systems filling each of the previous roles would
be specialized, and though by necessity highly
autonomous, one should not necessarily expect what
science-fiction gives us for each of these roles.
– A military robot may well not be at all like the
Terminator, a machine with distinctly military
objectives, but more-or-less human appearance and
diversity of abilities.
– A robotic sex slave may well not be at all like Gigolo
Joe, with the full diversity of human abilities
– A robotic servant may well be nothing like C3PO, a
‘droid programmed for a specific purpose but with
extremely impressive (ahem, humanlike) flexibility.
Specialization
• This specialization means that robots in
such roles aren’t likely to be totally
anthropomorphized, so their social impact
may be more difficult to consider or see.
Future regulation
• Wallach and Allen specify that they see no
concerns here so pressing that they
should be used to indefinitely arrest some
research projects, but they point out that
the time to examine the cultural impacts of
robots used in the previously mentioned
roles is before they are there, and while
they are being incorporated. (Again, think
“Engineering Activism”)
A Pervasive Criticism:
• There is a difference between a moral agent and a moral
subject, and generally, people are both. It is not difficult
to imagine something that is a moral subject but not a
moral agent (young children and animals can be said to
fill those roles). However, having something that is a
moral agent (something that owes us something morally)
but is not a moral subject (something we owe nothing to
morally) is a precise description of a slave morality.
• In referring to their goal as only the consideration of
AMA’s, the authors seem to, by omission of alternatives,
beg the question against the concerns addressed in the
first full paragraph of p. 50.