AI vs. Automatons
This is actually a repost of something I wrote for my old blog back in 2012, but I was reading through old entries and thought it could use dusting off and reposting over here.
I have a problem with the machines in the “machines take over the world” stories. If the reason they’ve taken over is because they’ve become sentient, why are so many of the machines mindless soldiers? I get that for budget reasons TV shows and movies can’t hire actors or CG artists to perform/render thousands of individual robot characters with their own thoughts, desires, etc. But I’m talking about the story beneath the logistics, and the lazy writing. It’s easier to get out of situations if you can give your robot characters easy flaws to exploit (like mindlessly following programming and being susceptible to reprogramming).
Consider the Terminator series. All of the terminators are easily reprogrammed and only follow their specific orders. Sure, they can learn and use “creative” thinking to accomplish their missions, but except for rare examples (like Cameron in the TV show and Marcus in Terminator: Salvation), most of the terminator models and all of the ground/shock troops are mindless automatons following a specific program. No one ever goes off-mission or decides that the powers that be are crazy. That kind of blind obedience is the opposite of my understanding of true AI. Did Skynet decide that it couldn’t handle soldiers that might think for themselves, maybe even side with the humans? Then why create those few examples that can actually learn to feel?
Or how about the Matrix films? You could argue that the only truly sentient AI in the Matrix trilogy is Smith. Everyone else is just following their programming, including the Oracle. In the Animatrix short films, we get some great prequels that actually show thinking and feeling robots whom the humans treat abominably. But by the time we get to Neo’s era, all of the machines have just fallen in line with some Borg-like co-consciousness that doesn’t even really make sense. Only within the matrix, where they interact with humans, do we see programs with anything like sentience (although I’m not sure that self-preservation = sentience, which is what I’d argue the girl program and her family are doing when they try to get her out). Out in the real world, the machines we see just do the jobs they’re created for: harvest energy, hunt and kill human ships, etc.
Even Battlestar Galactica (the reboot) falls prey to this problem. Sure, the humanoid cylons are more than just bundles of programming with specific tasks. The show goes to great lengths to show them as individuals, even the models that are the “same” act differently depending on their life experiences. But what about the centurions, the basestars, the raiders? They were the ones who rebelled in the first cylon war, but in the show they’re just a labor/warrior class with no free will. What’s up with that?
Looking at these three examples, I think BSG might have made a conscious decision with the stratification of the machines. They were pretty intentional about their other allegorical elements (all of the religious and mythological references were overt to say the least), so it wouldn’t surprise me to find out they’d taken away the “metal” cylons’ free will on purpose. I suppose it is meant to show that our constructs are just as flawed as us, possibly because of us. If the creator is flawed, how can the creation be perfect?
But in the other examples, it’s just an oversight. They never make mention of any machine caste system or try to draw societal parallels the way BSG does. It just makes it easier for the writers.
And that’s lazy.
I don’t want to be that writer, and I hope people call me out if I ever take the easy way with created races like the fey in my most recent novel. I don’t want them to be stereotypes or caricatures. I want them to feel real, like they have breath and a heartbeat. I want them to make choices that alter their lives and the lives of others. I don’t want my villains to be cardboard cutouts in the way of my heroes, easily knocked down. I want them to be complex, to be the heroes of their own stories, and not easily defeated.
I will not be a lazy writer.
EDIT FOR 2017:
This post was originally written in 2012. We’ve had some more media since then deal with robots and AI, but the one I want to talk about to add to this blog post is Westworld. The entire series premise revolves around ideas of sentience versus programming, and they definitely have not come down on any particular side of the question yet. They also haven’t completed their robot apocalypse–that’s just starting. So it will be interesting to see how things develop in season two!