AI is often sold to employees as a pure upgrade. It’s a way to write faster, analyze better, and perform at a higher level with less effort.
But John Nosta, an innovation theorist and founder of innovation and technology think tank Nosta Lab, said the framing overlooked a crucial downside: what happens after the boost.
In his view, AI does more than just improve performance. The absence of technology can also undermine the basic skills people rely on.
“The skill set is actually below the baseline,” Nosta told Business Insider, explaining what he calls the “AI rebound effect.”
AI makes you better, then it makes you worse
Nosta compared its effectiveness to doctors using AI to perform colonoscopies.
He said having AI scans alongside clinicians to help them find small polyps makes doctors better at their jobs. The problem arises when the same doctor performs the surgery the next day without the aid of AI, he said.
“We have to get back on track,” Nosta said. “And the skill set is actually below baseline.”
The danger, he said, is not just addiction, but regression.
As competence decreases, confidence increases.
Nosta also warned that AI could distort the way workers judge their abilities. It’s a concern shared by many academics and researchers, including Rebecca Hines, director of the Labor AI Institute, and Nobel Prize-winning physicist Saul Perlmutter, who say AI can impair judgment while giving the illusion of understanding.
“In fact, we are over-exaggerating our capabilities with AI,” Nosta said, calling the effects “really dangerous.”
In his view, AI does more than just help people do more. It makes them feel more capable, even if that confidence is not backed up by independent skills.
Once AI support is removed, that false confidence could be dangerous, especially in high-stakes environments, where employees may take on tasks and decisions beyond their actual judgment, he said.
Growing workplace risks
Nosta explained that he sees a growing “cognitive codependency” among younger workers, especially those in AI-infused jobs.
When used intentionally, he believes, AI can “make me smarter.” When used instead of thinking, he said, “you become stupid.”
Researchers at Oxford University Press came to a similar conclusion in a report published last October, saying that while AI will help students learn faster, it will reduce the depth of their thinking. Other scholars take it a step further.
Kimberley Hardcastle, a business and marketing professor at Northumbria University in the UK, told Business Insider last October that over-reliance on AI could lead to “atrophy of epistemic vigilance” – a shrinking of the ability to independently verify, challenge and build on knowledge without the aid of algorithms.
To avoid “cognitive atrophy,” “we need to maintain a certain level of cognitive risk,” Nosta said.
His prescription is deliberate resistance. This means maintaining “cognitive grit,” maintaining friction, and using AI to learn rather than avoid learning.
In the age of AI, he added, the biggest threat to jobs is not smarter machines, but humans gradually forgetting how to think without them.
“For the first time in history, human cognition is on the verge of obsolescence,” he said.
