Pull up a chatbot. Type a question. Get an answer. It's never been easier to skip the hard parts of thinking — and that ease might be costing us more than we realize.
As AI embeds itself deeper into everyday work, researchers are quietly sounding an alarm: heavy reliance on these tools is deskilling workers. Nearly half of employees surveyed by Workday last year feared AI would erode their critical thinking over time. It's not an unreasonable fear.
"Are we getting too dependent — and does that mean we're getting dumber as a result?" asked Anurag Dhingra, a senior vice president at Cisco. It's the question every major technological revolution eventually forces us to ask.
The answer, thankfully, isn't a simple yes.
Dr. Majid Fotuhi, a Johns Hopkins professor who studies neuroplasticity, draws a crucial distinction: passive reliance on AI weakens the mind, but active engagement with it can actually strengthen it. The difference lies entirely in how you use it.
So what does smart AI use actually look like? Experts across neuroscience, academia, and business point to a few core habits.
Go deep, not wide. Dhingra says investing real time in understanding a subject — not just skimming AI summaries of it — keeps critical thinking sharp. Gloria Mark, a professor at UC Irvine, echoes this: read long-form material, take courses that demand sustained focus, and resist "the path of least resistance." The goal is to stay in the loop, not opt out of it.
Write your draft first. Before opening any AI tool, get your own thinking down. Strategy leader Geetha Rajan always writes a first draft before using AI to pressure-test it. EY's Joe Depa tells his teams the same thing — draft it yourself, then use AI to refine and challenge. The point isn't to avoid AI; it's to make sure your ideas come first.
Make it argue with you. Vivienne Ming, chief scientist at the Possibility Institute, calls this "productive friction." Feed AI your argument and ask what you're missing, what the strongest countercase is, where you're wrong. That kind of friction is the opposite of intellectual laziness — it's a workout.
Slow down before you prompt. Most people use AI like a search engine, firing off questions and accepting whatever comes back. Jacob Sherson, a professor at Aarhus University, recommends a different approach: frame your problem carefully first, then use AI to generate multiple options, compare them, and actively critique each one. "If you only review one output," he says, "you are not evaluating — you are accepting."
Explain it back without looking. The clearest test of whether AI helped you think or just think for you? Close the screen and explain the reasoning out loud. Ming calls the failure mode here a "competence illusion" — AI-generated work that feels like your own, even when you couldn't reproduce it. If you can teach the idea, it's yours. If you can only recite it, it isn't.
The brain, like any muscle, weakens without use. The goal isn't to avoid AI — it's to make sure you're still the one doing the thinking.