Rocket Scientist

        One of many things my husband and I discuss are ideas that would work well in fiction.  Part of that is because my husband loves science and technology and anything vaguely animé-ish and part of it because we write fiction together.  Let me tell you, if you can stand to bring a creative concept to life, start to finish, with your significant other, you have the basis for a heckuva a relationship.  But I digress.

In any writing collaboration, there is the component that can, at least in general, spell.  That’s my part.  I take diverse ideas, some of which are mine, and weave them with characters I like for the story, then my husband listens to me read it out loud (I can’t promote the idea of reading aloud too strongly for any aspiring writer) and tells me what I did wrong.  Then we fix it.

He also provides some of the ideas.  Today, it was a discussion on AI, a particular favorite of his.  He was suggesting that, if one could have an AI, one could have them be raised as companions to children (in much the same way many animals/familiars/etc. have done in fiction) and learn their morals and judgement in a similar way.

That’s all well and good, I tell him, except I truly believe it takes more than environment that makes judgement and personality; I believe children are born with a personality, a spirit, a soul if you will.  It alone doesn’t shape a child into an adult; the environment is definitely a factor, but it’s the personality of the child that determines how a child responds to an environment.

So then the subject turned to whether this education would work on an AI that didn’t have an emotional response.  Lee seemed to be of the opinion that you couldn’t get to the learning level of intelligence without an emotional intelligence.  I’m not so sure. It’s not like a number of synapses automatically equates with emotionality.  The human brain is more complex than that and uses the amygdalae for much of this response (though not all).  How does one instill this in a machine?  And would you want to?

Teaching ethics and judgment requires a way of convincing children of what is right and wrong, not just what’s best for the child, usually using reward and/or punishment as an incentive or emotional empathy to help them understand the reasoning behind the decisions (”You wouldn’t want to be treated like that, would you?”).  But, if there is no emotional response to either, no opportunity for empathy, how effective can it be?

Of course, there is no right answer; we haven’t created sentience artificially (yet).  But the idea of each child growing up with a robot companion can definitely be of use.  I’m just not the kind of gal that can turn her back on an idea like that.  And I’m dangerous when I get an idea to worry.

Trackback URI | Comments RSS