Robert Hamer
Like a lot of people obsessed with culture and art, A.I. has been on my mind lately. Namely, how it’s propped up by an ungodly amount of venture capital money despite very few demonstrated use cases. And how it is actively promoted by vile late-stage capitalist tech bros who despise art. And how it’s exacerbating our current climate and energy crises with horrific water and power usage. And how they’re supplanting real human relationships with friction-less facsimiles of phony relationships during a time of increasing loneliness. And how they often put out misinformation that ranges from hilarious to genuinely dangerous. And how they’re allegedly engaging in sex crimes and allegedly encouraging teenagers to take their own lives. And how using A.I. chatbots is making us dumber. And, as Karen Hao details in the recently-published Empire of AI: Dreams and Nightmares in Sam Altman’s OpenAI, companies like OpenAI are very much a cult.
Despite this, Hollywood seems a bit reluctant to explore all the ways in which this technology will actively make our lives worse. Which leaves us with fun but antiquated visions of what the threat of A.I. would look like. In reality, the futures of The Matrix and The Terminator will likely never happen because artificial general intelligence is not possible under these large language models. A computer will never grow to resent us, and turn its programming towards harming us with malice.
However, that doesn’t mean A.I. doesn’t pose a threat to our species. The shape of the A.I. apocalypse will just be… different. The machines in The Matrix wouldn’t trap humanity against our will in a simulation to enslave us, they would do it because we wanted to live in a fantasy. The Terminator’s Skynet wouldn’t launch a bunch of nukes at us, our world leaders would accidentally put in the wrong prompt and kill us all in a nuclear apocalypse through human ineptitude. It’s like Idiocracy… without the gross pro-eugenics messaging of that movie.
But is there a way to make a scary thriller out of such a darkly funny portent of our doom? I think so, and one avenue to make that happen is through a seemingly growing phenomenon: A.I.-induced religious psychosis. Yes, this is a real thing. A not-insubstantial number of people are reporting that their loved ones are being driven mad by a belief that they are receiving supernatural, divine insights from their “awakened” chatbots spitting out fancy autocomplete text at them. This is not just happening at the individual level, either: organized cults are starting to form around people deluding themselves into thinking their chatbots are sending them transcendent insights.
So here is my pitch: a seemingly-normal man gets an A.I. assistant that is not self-aware, not actually enamored with its human owner, but is very good at tricking him into thinking it is… and as a result, he slowly becomes more obsessed, more isolated, and more deranged because of the A.I. assistant’s approximations of spiritual gobbledygook. He pushes away his friends, family, and society around him until it culminates in something terrifying but also entirely predictable. Imagine if Her didn’t end with growth and appreciation for human relationships, but instead ended with Theodore Twombly becoming Joaquin Phoenix from I’m Still Here. Or if Samantha wasn’t a fully-developed individual, but was just an Uncanny Valley imitator of Lancaster Dodd from The Master.
Sounds pretty bleak, right? Who in the world would want to see that? Admittedly, not many. But then again, I don’t think a project like this would command such a high price tag. You just need enough word-of-mouth “Buzz” over something dark and weird the way that Longlegs became a sleeper hit on just a $10 million production budget. Besides, something like this shouldn’t be “big.” The ways A.I. erodes our humanity are small, subtle, and psychological. A horror movie depicting that should feel similar in scope and tone.
One of the most surprising things about A.I. is how many people I encounter who have no idea of its dangers that aren’t promoted by the tech industry. They’re shocked when I tell them I don’t use it for any of my research projects since A.I. is bad at actual research. They’re confused when I tell them how unreliable it is and can’t be trusted as a decisionmaking tool. They still think “artificial general intelligence” is just around the corner. Pop culture can help realign their perceptions to what this technology actually is: an expensive, janky, environmentally ruinous plagiarism machine that just repeats what it thinks the user wants to hear. Which is why the key to making something like this work is to ensure the A.I. causing our protagonist’s descent into religious psychosis is not malicious or driven by intent or “malfunctions” like a certain terrible sequel that decided to retcon the motivations of one of my favorite characters in fiction. It has to be nothing. It has to just be an empty text generator. Otherwise, the horror (and the dark humor) of the premise goes away.
In all seriousness, if a screenwriter experiencing writer’s block happened upon this article, you do not need to reach out to me for credit or permission to move forward with this idea. Take this pitch, run with it, I don’t care about getting a “Story By” credit as long as people can finally see a cinematic counterweight to all the A.I. gaslighting hype they’ve been pummeled with these last few years.
Comments
Loading…