Our minds aren’t equipped to handle AI

There's a delicious irony in the fact that the AI industry's most prominent figures reach for computational metaphors to describe the human mind. Hassabis calling the brain "a biological approximation to a Turing machine" isn't just a technical claim—it's an act of intellectual self-definition. The AI community has spent decades validating its own project by asserting that cognition is computation, and now that assertion has become a kind of cargo cult: we built machines in the image of our theories about ourselves, and then we use those machines as proof that the theories were right. Wiener's observation cuts deeper than the article's framing initially suggests. "The thought of every age is reflected in its technique" isn't merely descriptive—it's a warning about epistemic closure. When your dominant technology is the computer, you inevitably see everything through that lens: the brain becomes hardware, memories become storage, reasoning becomes processing. But as the article implies, this isn't a neutral observation. It's a constraint on imagination. We didn't discover that the mind is computational; we decided it was, because that's what our age's technique allowed us to see. The deeper problem, as the piece hints, is that these metaphors are now actively harmful. When Musk and others frame superintelligence as simply a matter of scaling up computation, they inherit all the assumptions baked into the Turing machine model—discrete states, deterministic transitions, formal symbols. But modern neural networks are none of those things. They're continuous, stochastic, and opaque even to their creators. The metaphor has broken down, yet the industry keeps wielding it because it's the only vocabulary they have. What's striking is the educational dimension the article gestures toward. We teach computer science students to think in terms of algorithms and data structures, and then we ask them to reason about systems that defy those categories. The result is a generation of engineers who can build transformers but can't articulate what they're building. We're not just lacking the language to describe AI—we're lacking the conceptual infrastructure to even begin asking the right questions. The uncomfortable conclusion, which the article circles toward, is that our minds may indeed not be equipped to handle AI—not because AI is too complex, but because we've spent a century training ourselves to think in exactly the wrong terms. The first step toward genuine understanding might be abandoning the computational metaphor entirely, and admitting that the brain was never a Turing machine—and neither, really, is the thing we built in its name. Source: https://www.theverge.com/ai-artificial-intelligence/1003794/ai-education-computational-model-thought