AI is the new animal

Leo Kim makes the case for AI as a new companion species. This piece was originally published on January 12th, 2023.

There’s a new creature emerging from the primordial cybernetic soup of the 21st century. Fleshless, virtual, created in man’s image yet wholly inhuman. You’ve probably heard of its more famous incarnations: Dall-E, LaMDA, GPT-3. In short, the genus known as artificial intelligence—thinking machines. Even in these early forms, they’ve generated controversy for their uncanny encroach upon territories once thought to be the exclusive purview of homo sapiens (art, language, creative problem-solving, to name a few). It’s not quite SKYNET, or even Wall-E, but in these mechanisms, we see the outlines of an intelligence that threatens to unravel our given understanding of the technological and the sentient; something that forces us to ask, what kind of thing is this?

Some have heralded it as the next step in posthuman evolution; others, meanwhile, have fallen back onto theistic concepts: AI as the birth of God, or perhaps the devil. In most cases, there’s the shared view that these technologies represent some highly refined version of us—Humanity 2.0, now without the messy Humanness. And who could blame us for thinking this way? Philosophers for centuries have grounded human exceptionalism in things like language and creativity. Now, having created an intelligence that appears capable of outpacing us in these domains, it’s natural that we would see them as our more-perfect counterparts.

Yet having the human mind serve as the central point of reference through which we make sense of these intelligences leads us to ignore just how weird and alien they are, how little they actually resemble us. Though our longstanding belief in sentience as a distinctly human trait has caused us to relate to these intelligences as if we were alike—sparking endless debates about whether these computers are really better than us, if they make us obsolete—this anthropocentric analogy breaks down upon closer examination. Not only are AI often black boxes whose inner machinations are notoriously hard to crack, they frequently behave in ways that are inexplicable to us. Sometimes, it’s beautiful in its incomprehensibility, as AlphaGo was when it made a game-defining move against Lee Sedol that baffled commentators; other times, it’s horrifying, as was Google’s infamous algorithm when it struggled to make the basic distinction between images of Gorillas and darker-skinned subjects. 

In practice, this anthropomorphic projection isn’t just misleading, it’s dangerous, paving the way for the broader abdication of human responsibility. After all, if it really is the case that these technologies are better versions of us, then wouldn't they be more just, more knowing? This line of thought has already helped our technophilic fervor vest our new algorithmic philosopher-kings with the power to influence everything from parole sentencing to our tastes and desires, with disastrous consequences. 

Instead, what we need is a more nuanced way to relate to these other minds—to foster a relationship with these AIs grounded in the limits of their inhumanity and their unknowability. Luckily, this isn’t the first time that human technics have sculpted an intelligent, alien mind independent from our own. Today, we simply know these creatures as our dogs, cats, and furry friends: what Donna Haraway calls our “companion species.” These creatures not only instruct us on how we might live alongside thinking beings that possess a mind radically different from our own, but also reveal just how porous the lines between technology, humans, and animals can be. 

My childhood dog was, like me, a bit absentminded. She’d often zone out and lose the ball as we’d play in the backyard. As she ran around aimlessly, I’d point towards the neon-yellow Dunlop tuft sticking up from the grass. It always took her a while, but eventually, she’d notice me gesturing wildly and amble in that general direction. She’d find the ball, prance back—tail wagging, puppy dog eyes looking up imploringly—and we’d repeat the ritual. 

As modern humans, it’s easy to forget how astounding this seemingly unremarkable interaction is—how over the course of forgotten eons, two foreign forms of life became legible to each other through a series of technical innovations. The face of the dog, for example, changed so that it could display a range of emotions inarticulable to their wild ancestors (those puppy dog eyes we know so well). They learned to read us too, making sense of the hidden meaning behind our pointed fingers, our gestures and ever-changing expressions.

FURTHER READING