The comfort we get when offloading our cognitive load to LLMs is bad for us. Cognitive load should exist, and if we reduce it too much – if we stop thinking – we can actually unlearn how to think. Kids who always choose the easy route and copy their homework from other students eventually find themselves completely clueless about what’s going on in school. Someone who always lets their spouse handle all the bills and banking may one day be unable to manage even a simple payment on their own. A person who never bothers to learn street names or routes will be lost when their phone dies, not even knowing how to get home. It all comes back to what Nassim Taleb talks about with hormesis in his insightful book Antifragile – the idea that small doses of stress or discomfort make us stronger. Muscles grow by lifting weights. Immunity builds through exposure. Confidence grows by taking risks. Skills sharpen through repetition. Creativity expands by solving hard problems. The mind works the same way. The friction of thinking, the awkward struggle to find the right words – that’s mental weightlifting. Maybe a slightly different example, but of a similar spirit – the Broken Windows theory. The theory argues that visible signs of disorder – like graffiti, litter, or broken windows – signal that neglect is tolerated, lowering informal controls and inviting further misbehavior and even serious crime. In other words, when small cracks go unaddressed, they can cascade into much bigger fault lines. A constant reliance on LLMs will push this further and further until we’ve outsourced all our thinking, becoming little more than biological puppets. There’s even recent research that backs this up. In one study, participants were split into three groups: Brain-only: wrote essays without assistance. Search Engine: used Google search. LLM (ChatGPT): relied entirely on ChatGPT. And the results were: In the LLM group, 83% couldn’t quote anything from their own essays shortly after writing, whereas nearly everyone in the other groups could. Participants moving from LLM to writing solo exhibited reduced neural activity and continued under-engagement. Participants transitioning from brain-only to using LLM retained strong memory recall and showed neural activation patterns similar to the Search Engine group. Percentage of participants within each group who struggled to quote anything from their essays in Session They coined the term “cognitive debt” to describe the tradeoff: immediate convenience from AI assistance may come at the cost of long-term cognitive capabilities like critical thinking, memory retention, and creative autonomy. You borrow mental energy from the machine but with interest – and that cost shows up later when your own thinking has weakened. I keep saying that to my kids – LLM is a really powerful tool but use it wisely. Don’t ask it to solve your math equation; instead, provide your own solution and have it explain where you might be wrong. Starting with independent thinking and then integrating AI will be healthier for cognitive development. It’s like nuclear energy – you can use it for mass destruction or as a clean source of power. Consistent reliance on AI tools might undermine learning, memory, and creativity. Discomfort isn’t just a nuisance – it’s a training ground. Look for discomfort, seek it out, and encourage it.