12.12.2012 Views

Who Needs Emotions? The Brain Meets the Robot

Who Needs Emotions? The Brain Meets the Robot

Who Needs Emotions? The Brain Meets the Robot

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

EMOTION WITHOUT BIOLOGY<br />

beware <strong>the</strong> passionate robot 371<br />

Norbert Wiener, whose book Cybernetics: Or Control and Communication<br />

in <strong>the</strong> Animal and <strong>the</strong> Machine introduced <strong>the</strong> term cybernetics in <strong>the</strong> sense<br />

that is at <strong>the</strong> root of its modern usage, also wrote <strong>The</strong> Human Use of Human<br />

Beings (Wiener 1950, 1961). I remember that a professor friend of my parents,<br />

John Blatt, was shocked by <strong>the</strong> latter title; in his moral view, it was<br />

improper for one human being to “use” ano<strong>the</strong>r. I suspect that Wiener would<br />

have agreed, while noting that modern society does indeed see humans using<br />

o<strong>the</strong>rs in many disturbing ways and, thus, much needs to be done to improve<br />

<strong>the</strong> morality of human interactions. By contrast, robots and o<strong>the</strong>r<br />

machines are programmed for specific uses. We must thus distinguish two<br />

senses of autonomy relevant to our discussion but which often infect each<br />

o<strong>the</strong>r.<br />

When we talk of an “autonomous human,” <strong>the</strong> sense of autonomy<br />

is that of a person becoming a member of a society and, while<br />

working within certain constraints of that society and respecting<br />

many or all of its moral conventions, finding his or her own<br />

path in which work, play, personal relations, family, and so on<br />

can be chosen and balanced in a way that grows out of <strong>the</strong> subject’s<br />

experience ra<strong>the</strong>r than being imposed by o<strong>the</strong>rs.<br />

When we talk of an “autonomous machine,” <strong>the</strong> sense is of a machine<br />

that has considerable control over its sensory inputs and<br />

<strong>the</strong> ability to choose actions based on an adaptive set of criteria<br />

ra<strong>the</strong>r than too rigidly predesigned a program. 5<br />

On this account, a human slave is autonomous in <strong>the</strong> machine sense but<br />

not in <strong>the</strong> human sense. Some researchers on autonomous machines seem<br />

to speak as if such machines should be autonomous in <strong>the</strong> human sense.<br />

However, when we use computers and robots, it is with pragmatic humanset<br />

goals ra<strong>the</strong>r than “finding one’s own path in life” that we are truly concerned.<br />

In computer science, much effort has been expended on showing<br />

that programs meet <strong>the</strong>ir specifications without harmful side effects. Surely,<br />

with robots, too, our concerns will be <strong>the</strong> same. What makes this more challenging<br />

is that, in <strong>the</strong> future, <strong>the</strong> controllers for robots will reflect many<br />

generations of machine learning and of tweaking by genetic algorithms and<br />

be far removed from clean symbolic specifications. Yet, with all that, we as<br />

humans will demand warranties that <strong>the</strong> robots we buy will perform as stated<br />

by <strong>the</strong> supplier. It might be objected that “as adaptive robots learn new behaviors<br />

and new contexts for <strong>the</strong>m, it will be impossible for a supplier to<br />

issue such a guarantee.” However, this will not be acceptable in <strong>the</strong> marketplace.<br />

If, for example, one were to purchase a car that had adaptive circuitry

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!