"Artifical" Intelligence

(by Joe)

Watching the new Ridely Scott film Prometheus last night, I realised that there's something about the term “artificial intelligence” that doesn't quite sit right with me. SPOILER: there's an android (or humaniform robot, to use Isaac Asimov's term) in the film, one that for all intents and purposes behaves and appears like a human. A somewhat odd human perhaps, one that feigns a degree of subservience to those around it, but a human nonetheless. I would certainly be happy to say that it was conscious, and in terms of intelligence it far exceeds almost every other character in the film. However, I'm not so sure that I'm comfortable calling it “artificial”.

This is Idris Elba. He's not an android.

Early on in the film, one of my companions whispered something like “oh, so he's an AI then”, in what I can't help feeling was a slightly dismissive tone of voice. Whilst this is perhaps technically correct, or at least an accurate use of the term, I don't think that I'd have chosen to use it. Maybe I just read too much science fiction, or spend too long thinking about multiple realisability, but to label a conscious system “artificial” in this way seems distinctly discriminatory to me.

Of course, if like John Searle you think that a conscious, thinking robot is necessarily impossible, then this won't bother you very much. I'm not going to argue for the possibility of Strong AI here, but suffice to say I am essentially a functionalist about consciousness, and thus am firmly committed to the possibility of conscious awareness being instantiated in a non-biological system. Such a system, if we had built it, would be “artificial” in the sense that it would be a constructed artefact, but to label it as such risks distorting our understanding of what it actually is. Referring to an intelligent android as an AI distances it from ourselves, putting it in the same conceptual category as a mindless computer or microwave. We would be tempted to treat such a creature as no more than a tool, and there is certainly an air of dominance towards our creations that the term “AI” can only help reinforce.

In fact, the film managed to address this issue. One otherwise very empathetic member of the ship's crew behaves in a distinctly abusive way towards the android, making constant remarks about how inhuman it is, and treating it as little more than a slave. This behaviour was reminiscent (purposefully, I think) of colonial attitudes towards indigenous populations, being patronising, cruel and dehumanising. I don't think that it would be unreasonable to say that this character was being “racist” towards the android, although we perhaps need a new word for this particular form of discrimination. “Instantialist” is somewhat clumsy, but it gets the point across. I believe we will, in the relatively near future, develop computer “minds” that are functionally similar enough to be thought of as conscious, and when this happens we will be faced with an ethical dilemma. Should we be allowed to treat these creations as creations, or should they be afforded just as much dignity and respect as any other intelligent life-form? We risk inventing a whole new category of discrimination, one that I believe the term AI, with all its connotations of subservience and inferiority, will only exacerbate. 

(The film, by the way, is well worth seeing!)