Thinking by Proxy: Why AI is Seducing Us Into Stupidity

Whenever anyone frets about AI turning us into idiots, they usually sound like Luddites pointing towards calculators or GPS. This reminds me of Frank Herbert’s sci-fi epic Dune where he imagined a universe where humanity became so slavishly dependent on “thinking machines” that their minds began to atrophy, eventually sparking a galactic crusade to ban computers entirely.

In the real world, of course, calculators didn’t ruin maths, and Waze didn’t destroy our ability to drive. In fact, these apps are classic examples of cognitive offloading: the sensible habit of outsourcing tedious mental chores to technology.

Most of the time, the trade-off is a bargain. We get to solve complex equations without sweating the long division, and we navigate rush-hour traffic without having to wrestle with a paper map.

What is the hidden cost of letting machines do the work?

There is a catch. If you deliberately rig a calculator to spit out obviously absurd answers, a shocking number of university students will simply write them down without batting an eyelid.

Rely on Waze every day, and your brain’s natural sense of direction begins to fade; studies show that drivers glued to their screens suffer a genuine drop in spatial memory. The internet did the same thing to our recall. Once we know we can just look up a fact online, our grey matter no longer bothers trying to retain it.

How does AI turn helpful offloading into mindless surrender?

AI takes this familiar problem and makes it much worse. The core commandment of the Dune universe is famously:

“Thou shalt not make a machine in the likeness of a human mind.”

Ironically, that’s literally the description of a neural network, one of the many algorithms used to build genAI models. Unlike a dumb calculator, a modern chatbot speaks with the smooth, persuasive confidence of an expert, even when it is making things up out of its ass.

Handing over basic tasks is one thing, but AI’s charming, conversational interface invites something much riskier: cognitive surrender. When researchers at the Wharton School tested volunteers by feeding them chatbots that randomly slipped false information into their answers, the humans simply stopped thinking for themselves. Rather than spotting the mistakes, they swallowed the bad advice whole.

How can bosses stop their teams from switching off their brains?

Right now, most employers are so desperate to get staff using AI that they aren’t worrying about the brain rot that might follow. Yet businesses still desperately need people who can think critically: not least because these models still make embarrassing blunders that require a human eye to catch.

So, what’s the fix? While we don’t need to launch a holy war against computers and train human “Mentats” to do all our calculations from scratch, managers can push back against mindless automation by introducing three pragmatic safeguards:

·         Hire people who enjoy thinking: Those with a natural appetite for intellectual heavy lifting are far less likely to blindly defer to a bot.

·         Incentivise scepticism: Offering real rewards and immediate feedback for spotting errors naturally wakes up our innate critical instincts and encourages users to challenge the bot.

·         Enforce digital sobriety: Build in regular periods where staff have to work without digital training wheels. Researchers recently found that chess students who only received AI tips at restricted moments improved their game twice as fast as those who could tap a button for advice whenever they wanted.

Outsourcing the drudgery is fine; surrendering your judgement is another matter entirely.