The tech bros are at it again, clutching their algorithmic pearls like crystal balls and insisting with all the modesty of a rooster at dawn, that their silicon savant will soon outthink humanity—and look better doing it.
They call it Artificial General Intelligence, or AGI—machines as clever as us, though blessedly lacking our taste for sarcasm. All men like Sam Altman and Eric Schmidt require of you is a few hundred billion of your dollars and the kind of indifference to environmental ruin that would make even a locust blush. Yet, much like their other proclamations, this one too deserves a raised eyebrow and a stiff drink.
Reality, with its knack for comedy, prefers to trip these fantasies on their grandiose faces. Professor Iris van Rooij and her team from Radboud University deliver the punchline with flair: AGI isn’t just improbable—it’s computationally impossible. Their research, published in Computational Brain & Behavior, pops the AGI bubble with mathematical precision, exposing it as little more than a pipe dream dressed up in a lab coat.
In an interview with Radboud University’s research newsletter Professor van Rooij doesn’t mince words when addressing the fervor surrounding AGI. “Some argue that AGI is possible in principle, that it’s only a matter of time before we have computers that can think like humans think. But principle isn’t enough to make it actually doable. Our paper explains why chasing this goal is a fool’s errand, and a waste of humanity’s resources.”
She told the newsletter, "If you have a conversation with someone, you might recall something you said fifteen minutes before. Or a year before. Or that someone else explained to you half your life ago. Any such knowledge might be crucial to advancing the conversation you’re having. People do that seamlessly."
He co-author Olivia Guest punctures the current Techbro-hype with icepick cool, "There will never be enough computing power to create AGI using machine learning that can do the same, because we’d run out of natural resources long before we'd even get close."
Their proof, elegant in its mathematical rigour, exposes a fundamental problem: the computational complexity of replicating human cognition. While today’s tech titans are busy showcasing AI systems that dazzle at parlour tricks like mimicking human speech, van Rooij’s team shows that such feats are a far cry from the domain-general intelligence that defines humanity. As the researchers aptly note, achieving AGI isn’t like climbing a steep hill—it’s like scaling Mount Everest with a pogo stick.
The core issue lies in the concept of "intractability." In plain terms, even with all the computing power in the universe, training an AI to truly mimic the generality and nuance of human cognition would require resources so astronomical that we’d run out of stars long before success. This isn’t a limitation of today's technology—it’s baked into the very fabric of the problem. AGI is less a moonshot and more a mirage.
Yet the fantasy persists, fuelled by a potent mix of hubris and marketing. Van Rooij’s team likens these so-called advances to decoys—clever illusions that distract us from the profound limitations of AI. Just as a magician uses sleight of hand to conjure impossible feats, AI companies conjure claims of AGI inevitability to conjure venture capital. The effect is enchanting; the implications, less so.
Because here lies the danger: these illusions don’t just mislead investors; they distort our understanding of ourselves. If we believe AI can match human cognition, we risk mistaking its outputs—statistical guesses at best—for genuine thought. This isn’t just a theoretical problem. As van Rooij warns, confusing “maps for territory” leads to poorer models of human cognition, undermining both science and society.
But there’s hope in a pivot. Instead of tilting at the AGI windmill, van Rooij proposes reclaiming AI as a theoretical tool for cognitive science. By using AI to test hypotheses about human thought rather than to replace it, we might actually advance our understanding of the brain—without the quixotic quest for mechanical minds.
Let the dreamers chase their impossible dream. Meanwhile van Rooij and Guest seem to be saying, “The rest of us have work to do.”
