Have you read The Lord of the Rings^[1] by J.R.R. Tolkien? 

Or, even, seen the films?

Look, if you haven’t stop whatever you’re doing right now… and go read them. You will not regret it. Then come back and continue reading this post, living your life, existing in the world.

Okay? We’re on the same page? Good. Because I recently have seen a couple people mention an interesting theory—I’m not clear on the originator—about the titular ring in the story. The One Ring, the all-powerful more-than-a-macguffin of the story has often been cited as a stand-in for an all powerful and destructive weapon. For years we’ve been thinking about it as an allegory for nuclear armaments. But—and so the theory goes—what if a better (read: more prescient) analogy maps to the concept? What if the One Ring is AI?

So the theory goes…

In the story nearly everyone who encounters the One Ring believes with absolute certainty that they are the one who can wield it for good—but ultimately they are deceived and cannot. It corrupts every last one, even Frodo the ring-bearer chosen for his purity, strength and innocence, eventually succumbs. The ring, after all, has it’s own intentions and desires and is merely using the bearer to serve a secret darker master.

Map the determination of any commercial user of AI with the intention to do good with the tool to the ultimate outcome: creative shifts, slop generation, intellectual laziness, cheating in school, offloading work and feeding a hidden corporate backend with unclear motivations. Hmm.

The One Ring atrophies everything about the one who carries it, and we are given the creature Gollum as an example of what happens to someone who holds the powerful artifact for centuries: he is narcissistic, lacking empathy, and devoid of his remaining “humanity” (for lack of a better word.) The One Ring after all, tells the wearer everything they want to hear and acts as a kind of whispering sycophant validating every thought and notion.

Cognitive offloading and atrophy of human skills—or failing to learn the skills in the first place—is one of those fears often levelled at the shiny marketing pitches for AI. All the while, the LLMs chat sweet affirmations that every notion we have is brilliant or a thoughtful twist on an old idea as they whisper with certainty that we are bright and intelligent and nigh infallible. 

The creation of the One Ring is explained more deeply in the supplemental texts of the novels, but we are given to understand that it was forged in the fires of a mountain so hot and evil that it now lays at the heart of the enemy’s territory. Mount Doom was where the ring was made and the only place it could be unmade, created in darkness and trickery with immoral purpose enshrouding its entire history.

AI, famously, has been trained on stolen intellectual property. It consumes huge quantities of energy in data centres whose presence devalues communities and fills the air with noise and pollution, and the stains the water with heat and pollution. Talk about an allegory for Mount Doom.

Look, literature is how we understand the world, in particular science fiction and fantasy which map modern ideas onto unfamiliar worlds—it is to help us overcome our bias and see the same problem from a different perspective. Tolkien probably never saw a computer, let alone a multi-core GPU running a neural network model, but he saw the dark capabilities of human nature. He understood what happens when we etch shadows of that nature onto objects and systems outside of it. And that’s real the allegory, I think.