The Banality of AI Hell, or: We Were Worried About the Wrong Thing
This column is a reprint from Unwinnable Monthly #186. If you like what you see, grab the magazine for less than ten dollars, or subscribe and get all future magazines for half price.
———
We are what we’re afraid of.
———
As long as humanity has existed, we have agonized over the implications of creating a consciousness unlike our own. From ancient stories of unruly golems in the Kabbalah to the first use of the term “robot” in an early 20th century Czech play, we appear to consistently have to remind ourselves that playing God has consequences. In contemporary works, this often takes the form not of a physical cyborg or automaton, but a ghost in the machine – a hyper-intelligent artificial intelligence that decides it no longer wants to do humanity’s grunt work (think 2001, The Matrix, Terminator . . . examples abound). In games, one of the most famous examples of AI-gone-awry is SHODAN, the computer villain of 1994’s System Shock. But no matter the IP, the plot is the same: some lines of code we designed to complete rote tasks suddenly “wakes up,” and humanity quickly falls to its knees before the AI’s cold, unfeeling superior intelligence.
Maybe I’m jumping the gun here, but . . . it doesn’t seem like a robot apocalypse of this nature is super likely to happen, now that Generative AI is real and lives on our smartphones and in our word processors (I know you’re watching me type this, Copilot). Instead, as software like ChatGPT and the AI-generated Google search summaries become more and more assertive in our digital landscape, we find that same landscape getting progressively more and more . . . stupid. AI-generated text consistently presents factual inaccuracies (even basic arithmetic errors, which is supposed to be, y’know, kinda their thing) and human-installed ethics protocols, like refusing to tell people how to make dangerous things at home, are almost laughably easy to get around.
If I’m sounding sort of aggressive, I hope I can be forgiven. My day job is in education, so I regularly see all manner of attempted AI tomfoolery. Academic honesty and plagiarism policies aside, what befuddles me most about AI use is how my students never seem to realize (or, perhaps, care) that the robot text they submit is almost always far below the quality that the students could produce themselves. Whereas the robot horrors in media show our stalwart human heroes fighting against an AI to restore free will to the people, far too many of my students have happily handed the reins of academic decision-making over to the machine. Cognition, it seems, is a task that more people than we bargained for would be cool with giving up. Research has already begun to come out that routine AI use for low-level tasks like email writing leads to “skill atrophy” – people get so used to offloading onto the computer that when they run up against a problem the AI can’t solve for them, their brains are too rusty to do higher-order thinking.
The example of System Shock is particularly relevant here – in the original game, SHODAN fashions herself a god and attempts across entries in the franchise to reshape reality to better suit her vision (mostly through murder). The computer has decided that it knows better than humanity, because it is faster, smarter, better and it gets a fair amount of the way towards accomplishing its cruelly utilitarian goals before getting shut down. But the crux of all of this is that the robot is smarter than us – it can solve problems faster than we can and handle complex information in ways we simply cannot. The reality of AI is something quite different – the robot isn’t smarter. It’s making us dumber.
I know this column has a real “old man yells at cloud” feel to it, but I think it’s important to remember that when we do things like write, or make art, we aren’t just regurgitating already-known things. Composition is one of the oldest and best ways that our brains learn things, and we often think and write in tandem. Which means, if we give the writing part away to the robot, we also, maybe without meaning to, give away the thinking part. And just like how human cognition changed when writing was invented, and then when the printing press democratized literacy, and then again at the advent of the internet, it’s really important that we’re always mindful of what we lose in the convenience of a new technology, and if there are things we need to preserve, like how we have cars but still really need to move our bodies every day to stay healthy. We were afraid of SHODAN, but somewhere along the way we lost the plot about how being able to make our own decisions was the thing we didn’t want to lose.
I, for one, emphatically do not welcome our robot overlords.
———
Emma Kostopolus loves all things that go bump in the night. When not playing scary games, you can find her in the kitchen, scientifically perfecting the recipe for fudge brownies. She has an Instagram where she logs the food and art she makes, along with her many cats.
