(Image credit: Netflix)
I use AI every day for work. I ask ChatGPT for everything from help to shifting my productivity mindset to parenting advice. I use Gemini all the time, especially NotebookLM, and I regularly use Claude to code. In other words, I'm far from anti-AI.But when comedian Ronny Chieng recently told Harvard graduates that “AI is gonna end up making mediocre people dumber,” the line immediately landed.It’s harsh, but it gets at a feeling many regular AI users have but haven’t said out loud. The real concern isn't just that AI will take our jobs; it’s that it will quietly take over the parts of our brains we used to rely on — writing, remembering, synthesizing and sitting with a problem long enough to figure out what we actually think.The danger of mental autopilot
Ronny Chieng Address | Harvard Class Day 2026 - YouTube
Watch On
A recent TIME report highlighted emerging research into whether a heavy reliance on AI could weaken cognitive skills. But the answer isn't as simple as "AI makes you dumb."The better question is: Are we using AI to think better, or to avoid thinking at all? Chieng’s Harvard speech was an exaggeration meant to wake up a crowd. He told graduates to “destroy AI,” but underneath the joke was a real warning: Don’t let AI steal the satisfying, messy, human part of work.Anyone who has used a chatbot long enough knows the feeling. You ask for project topics, pick the least-bad one and move on. You ask for an email, skim it, and hit send. You ask for a summary, absorb the bullet points and never read the source material.Get instant access to breaking news, the hottest reviews, great deals and helpful tips.That may feel highly efficient, but if it becomes your default mode, you stop practicing the difficult part of thinking.Tools have always changed how our minds work. Calculators changed math; Google changed memory because we can look up anything instantly; GPS apps like Google Maps changed navigation. None of these tools automatically made humanity stupider, but they did create trade-offs.For instance, if you rely on Google Maps constantly, you get to your destination, but your internal map never improves. AI does the same thing to your brain. You can ask ChatGPT to write the memo, summarize the meeting and organize the plan. The box gets checked. But did you get any better?Passive vs. active users: The real AI divideThe danger is not that AI exists. The danger is turning every hard mental task into a prompt. The best AI users aren't using the best prompts; they are the ones bringing the most judgment. That is, their own human judgment.There is a massive difference between asking AI to write something for you and asking it to challenge what you already wrote. There is a difference between using ChatGPT as a ghostwriter and using it as a sparring partner.I get the most out of AI every time I use the tools to help me pressure-test an idea, find the weak spots or push me to explain what I really mean. Conversely, if you open a blank chat and ask it to "write something about [topic]," the result is always generic garbage.AI amplifies what you give it. If you give it curiosity and judgment, it helps you go further. If you give it laziness, it produces polished laziness. For that reason, AI creates two distinct types of users:Passive users: Ask AI to do the thing and accept the result. They outsource.Active users: Ask AI to help them understand the thing or see it from another angle. They collaborate.AI may not make people dumber because the technology itself is harmful. It makes people dumber when they use it to stay mediocre and do less.How to use AI better (and stay sharp)The answer isn't to stop using ChatGPT, Gemini or Claude. The answer is to stop treating AI like a replacement for effort.Here is how to stay an active user:Answer first: Before asking ChatGPT for ideas, write down three of your own. Before asking it to summarize an article, skim it yourself. Force yourself to have an opinion first, then bring in the AI.Use it as a critic: Instead of asking "Write this for me," try prompts like: "What am I missing here?" "Where is this argument weak?" or "What would a smart skeptic say about this?"Treat it as a blunt assistant: Use it to pressure-test your logic, not to bypass the act of thinking entirely.Bottom line: useAI intelligently The risk of using AI is that it makes it incredibly easy to stop trying. Used badly, AI turns thinking into a transactional button-press. Used well, it forces you to question your assumptions and refine your ideas. The future won't belong to people who use AI or those who avoid it — it will belong to the people who still know how to think for themselves.
