What's particularly troubling is that this is happening fast. ChatGPT only launched in late 2022. We're less than four years in, and professors are already describing fundamental shifts in how students approach learning. We're running an experiment on an entire generation, and we won't know the results for decades.

The cynical take is that this doesn't matter—if AI can do the work, why should humans learn to do it? But that logic breaks down when you consider that AI doesn't understand; it pattern-matches. It can regurgitate information and remix it in novel ways, but it can't think critically or question assumptions or create genuinely new ideas.

Unless we're willing to bet that AI will always be smarter than humans and available at all times, we probably still need humans who can think. The question is how we teach that skill when the default mode is to ask an AI.

Some professors are adapting by making assignments AI-proof: real-world projects, oral presentations, reflective essays about personal experiences. But that only works at small scale. You can't give personalized, un-Googleable assignments to 300 students in an intro course. The economics of higher education don't support that level of customization.

Others are embracing AI and teaching students how to use it effectively. The logic is sound: if AI is going to be part of their professional lives, they should learn to use it in school. But there's a difference between "use AI as a tool" and "let AI do all your thinking." We haven't figured out how to teach that distinction.

What keeps me up at night is the compounding effect. Students who use AI through high school and college enter the workforce without ever having struggled through hard problems on their own. They become managers who've never learned to think critically. They make decisions based on AI suggestions they can't evaluate. The degradation is gradual but irreversible.

The technology is impressive—ChatGPT and its competitors really can write coherent essays and solve math problems. But the question isn't whether the technology works. The question is whether we're okay with a generation that never learned to think for themselves because a machine was always there to do it for them.

Professors are fighting a losing battle, and they know it. You can't compete with a tool that's available 24/7, never gets tired, and gives instant feedback. But maybe losing this battle means we need to rethink what education is for. If the goal is to transfer information, AI wins. If the goal is to teach people how to think, we need to figure out how to do that in a world where thinking is optional.

5 days ago