Thursday, May 08, 2014

Zombie as Automaton

The zombie as a metaphor for the fear of immigrants is compelling and convenient.  As such, it holds much explanatory power, especially given the modern popularity of zombies.  It is not the only metaphor available, however, and it may not even be the best modern metaphor.  What follows is a brief meditation on zombies as a metaphor for the fear of automation.

Popular conceptions of zombies have always been relentlessly “other”, easily distinguishable from normal humans by their shambling gaits and apparent mindlessness.  If we relied only on the portrayal as other, the immigrant metaphor would suffice.  Yes, there are easy racial overtones, in the same way racist propaganda has sought to dehumanize various peoples in some way, especially in the focus on skin color as a means of determining the degree of otherness, but also in the various attempts over the years to compare intelligence scores between racial groups in an effort to support human-animal classifications that placed non-white races as more akin to animals.  And let’s face it, most zombies do act like “animals” in some sense.  I suspect that’s why these metaphors persist.

But what if, instead, we focus only on the apparent sense of self, that quality we like to use to raise humanity (all of it, if we’re feeling humanist enough) above its mere animal origins?  Sure, there is the animal connection again, the human-biased denial of the selfhood of animals.  As science slowly chips away at the supports of that argument, however, the human animal is left grasping at something else above which to raise itself.  Automation, specifically artificial intelligence, seems to be the target.  This is where a concept of self as a uniquely human (or, as we admit broader definitions of sentience, animal) trait begins to converge into a fear of artificial intelligence.  Even the name artificial intelligence implies a hierarchy, with natural intelligence trumping artificial.

In his novel Blindsight, Peter Watts paints a picture of a zombie that is so good at blending in it doesn’t know it is a zombie.  Possibly of greater importance is that neither does anyone else.  It has adapted itself to act like other people, and although it has its quirks, it appears to fall along the spectrum of roughly normal human experience.  If you are afraid of the implications of artificial intelligence, then this is the zombie that will frighten you the most.  Our intuitive understanding of such artifice is that, no matter how convincingly it portrays itself, it cannot possess a “true” sense of self.  

If we move back into the realm of popular portrayal, the metamorphosis from human to zombie requires the death of the self.  In almost no portrayals do zombies have memories of their former lives, and they instill terror by imbuing non-human life into the dead masks of people we knew.  The most dangerous people in zombie movies fall into two groups: 1) those infected who were, in life, closest to the protagonists, and 2) those infected who gain access to the protagonists’ inner circle prior to becoming recognizably other.  Both are predictable but crucial milestones.  The first requires the protagonist to confront the death of self in former loved ones, thus allowing the protagonist a viable psychology of survival against all other zombies (i.e., they become inhuman).  The second occurs later and forces the protagonist to implement a refinement of the psychology by which tests of worthiness can be applied to any applicant to group membership.  

Both scenarios can appear in a fear of automation metaphor.  We live in an age where it’s no longer readily apparent which actors with which we deal are real humans, pure automaton, or something in between.  Our world is populated by a chimera army of mechanical turks and weak artificial intelligences who, while perhaps unable to pass a Turing test, are nevertheless convincing enough that at worst, they drag us kicking and screaming into the uncanny valley but at best leave us guessing.  Add to this the existential dread conjured up by the specter of genetic engineering, mind uploading, and the perennial efforts to achieve artificial intelligence, and you have a recipe for horror that cannot easily be mapped to a horrific portrayal.  And people have tried.  It’s hard to be truly afraid of a robot uprising because there is rarely a portrayal in which the robots themselves aren’t recognizably other.  Notable exceptions include Terminator, which provides some of the most compelling horror precisely because the terminators resist identification as other.  Identification as other is the prerequisite for being able to fight against the other.  

Zombies that are more successful at appearing human will provide the basis for the most terror, according to this idea.  And the zombies that are most likely to be successful at appearing human are the automatons we’ve either created or made of ourselves, since they reduce our sense of recognizing other-than-human actors.  We fear (either for ourselves or on behalf of our loved ones) being stripped of the illusion of consciousness and having it replaced by a convincing proxy that we cannot fight because we cannot recognize it.