A lot of people, especially in younger generations, seem to be getting less intelligent. I notice this in myself. When I use ChatGPT to do tasks I used to handle on my own, I feel my own cognitive ability slipping. I no longer read long research papers in full — I paste them into ChatGPT. Many other college students, and people both older and younger than me, do the same.
There are three main reasons for this decline in knowledge, comprehension, and critical thinking.
The first is the rise of short-form video content. YouTube, TikTok, and Instagram have pulled people away from reading and reflective thinking. Dopamine-driven content keeps us constantly occupied, eliminating the downtime where people used to read, daydream, or simply think. Reading is probably the most effective way a person can build a mental model of the world — it forces you to connect ideas, follow arguments, and categorize new information. Short-form content does the opposite. It exposes you to small, fragmented topics driven by an algorithm you cannot meaningfully direct, making it nearly impossible to develop coherent, lasting knowledge.
The second is COVID-19. During the pandemic, many young students had years of foundational learning and intellectual discipline disrupted. With schools closed or operating remotely, an entire generation missed the structured environment where reading, writing, and sustained concentration are built. These are not skills that catch up easily. The effects are long-lasting, and we are still early in seeing how they play out.
The third is cognitive offloading through AI. More and more, people rely on AI to think, summarize, and decide for them. The divide here is already visible: elites use AI as a tool to amplify their existing intelligence, while most people use it as a substitute for thinking, gradually weakening their ability to process complex information independently.
It is already documented that people are reading less than before. Fewer people read the news in any serious form, and those who do often rely on summaries, headlines, or algorithmically curated clips. Comprehension skills deteriorate without practice. This is creating a widening gap — the masses are getting less capable, while a smaller group of people, equipped with AI and still possessing the foundational skills to use it well, are accelerating away from everyone else.
Thomas Piketty argues in Capital in the Twenty-First Century that capital grows faster than economic output, which compounds the wealth gap over time. The intelligence gap works the same way. The people best positioned to use AI are those who already think clearly, read widely, and can evaluate what AI produces. That advantage compounds. The result could be a new kind of aristocracy, not just of money but of cognitive capacity, with big tech companies already controlling what information most people encounter daily.
The political consequences are just as serious. Democracy depends on a citizenry capable of critical thought. Without it, people are more easily swayed by propaganda, misinformation, and polarization. We are already seeing this. If the trend continues, the conditions for genuine self-governance erode.
AI does not have to accelerate this. Used well, it could expand human intelligence rather than replace it. The problem is that most people do not know how to use AI to research, stress-test ideas, or build real understanding — they use it to avoid doing those things. Better education, better design of AI tools, and deliberate policies around how AI is integrated into learning could reverse the dynamic. The technology exists to make people meaningfully smarter. Whether that happens depends on choices being made right now.