The Silent Pandemic: How AI Is Eroding Our Ability to Think

Written by David McMahon

We are training our brains for dependency, sacrificing critical reasoning for the illusion of efficiency.

A quiet pandemic is spreading, not through our bodies, but through our minds. It’s a plague of cognitive erosion, a slow-burn deskilling of the human intellect, and its vector is the artificial intelligence we’ve so eagerly welcomed into our lives. We are becoming a society that generates first and thinks later, a world where the friction required for genuine reasoning is being smoothed away by the seductive ease of AI. The result is an entire generation at risk of losing the very skills that define human intelligence: critical thought, independent reasoning, and deep learning.

This isn’t hyperbole; it’s an evidence-based crisis. A landmark 2025 study from the MIT Media Lab, titled “Your Brain on ChatGPT,” used EEG scans to peer into the minds of students writing essays. The results were stark. The brains of those using an LLM showed the “weakest overall coupling” and “consistently underperformed at neural, linguistic, and behavioral levels” compared to those using a search engine or no tools at all . Researchers described the phenomenon as accumulating a “cognitive debt”—a measurable decrease in learning skills. Most damningly, the majority of students who used AI couldn’t even quote a single sentence from the essays they had just finished writing . They had become mere conduits for generated text, not authors of their own thoughts.

This mental outsourcing has a name: cognitive offloading. As cognitive scientist Misia Temler explains, we constantly balance offloading (letting others think for us) with scaffolding (using external sources to enrich our own thinking) . But with AI, we are not scaffolding; we are systematically dismantling the structures of our own minds. A 2025 study in the journal Societies involving 666 participants found a significant negative correlation between frequent AI use and critical thinking ability, a decline directly mediated by this increased cognitive offloading .

The neurological basis for this deskilling is becoming alarmingly clear. Research published in Frontiers in Medicine explains that as we repeatedly offload tasks, our brains adapt. The prefrontal cortex, the seat of planning and problem-solving, becomes less active. The hippocampus, crucial for memory, shows reduced involvement. Our brains, wired for efficiency, are being rewired for dependency. Dopaminergic reward systems reinforce the use of these easy, AI-driven strategies, making us neurologically more likely to favor cognitive offloading over effortful thinking . In essence, our minds are shifting from flexible, analytical networks to automatic, habit-based circuits. We are training our brains to be lazy.

“The mind is a muscle like any other. When you don’t use it […] that muscle atrophies incredibly fast,” one young adult lamented in a 2026 Harvard Business Review survey. Another stated, “Chatbots allow you to access information, not process it” .

This isn’t a new anxiety. Nicholas Carr warned of this in his 2010 book The Shallows, arguing the internet was already promoting “cursory reading, hurried and distracted thinking, and superficial learning” . He later explored the concept of “automation complacency” in The Glass Cage, using the example of airline pilots who, over-reliant on autopilot, lose their intuitive flying skills . We are now witnessing this on a societal scale. From medicine, where clinicians risk losing independent diagnostic reasoning , to the knowledge economy, where the fundamental skills of writing, coding, and analysis are being outsourced to algorithms, the deskilling is pervasive.

Some argue that AI is just another tool, like the calculator or GPS, and that society will adapt. But this comparison is a false equivalence. A calculator automates a single, rote function; generative AI automates the very process of synthesis, of creation, of thought itself. As Harvard lecturer Dan Levy notes, “If a student uses AI to do the work for them, rather than to do the work with them, there’s not going to be much learning” . The critical distinction is whether we use these tools for scaffolding—to enrich and support our own thinking—or for offloading, to replace it entirely.

The evidence overwhelmingly suggests we are choosing the latter. A generation is now being raised with the default behavior to generate first and think later. Neuroscientist Jared Cooney Horvath has warned that Gen Z may be the first generation to be less cognitively capable than their parents, citing declining literacy and numeracy scores over the past decade . The friction of difficult, effortful learning is what makes it deep and transferable. By removing that friction, we are not making learning easier; we are making it shallower and, ultimately, less effective.

We stand at a precipice. The convenience of AI is undeniable, but the cost is becoming terrifyingly apparent. We are trading the invaluable, hard-won skill of critical reasoning for the fleeting ease of instant generation. If we do not consciously change course—in our schools, our workplaces, and our personal habits—we risk becoming a society of intellectually passive consumers, brilliant at prompting but bankrupt of original thought. The silent pandemic will have run its course, leaving behind a world where we have forgotten how to think for ourselves.