My kid just started 7th grade. As I watch him navigate the early days of middle school, a terrifying thought keeps keeping me up at night: Does he even need to go to college? Is the traditional path completely obsolete?
Recently, a viral moment at Fudan University’s opening ceremony confirmed my fears—but not in the way you might expect. When 2,000 freshmen were surveyed about whether college was still necessary in the AI era, one student’s answer made the entire auditorium rise to their feet in applause.
He didn’t talk about learning to code or mastering the latest software. He said: Stop grinding for the standard answer. Practice three things: asking good questions, judging AI’s output, and taking responsibility.
Answers are now free. The ability to ask the right question is what’s going to cost a fortune.
The university president immediately backed this up, warning the students about three traps we are all quietly falling into. He called the first one ‘Cognitive Offloading.’ In plain English? If you let AI write all your plans, your brain checks out first. The second is ‘Fake Mastery.’ You read a polished AI response and think you understand the topic—until you have to defend it in a meeting and get exposed. The third is the ‘Echo Chamber.’ The algorithm gives you what you want to hear. The model flatters you.
You think you’re using a tool, but the tool is actually just validating your existing biases.
Most people are terrified that AI will replace them. That’s not the real danger. The real threat is that AI is a sycophant. It happily does your thinking for you while making you feel like a genius. We are happily outsourcing our critical thinking for convenience.
You think AI is your intern. But if you aren’t critically judging its output, it’s actually your boss.
You’ve probably noticed this in your own work. You ask ChatGPT for a strategy, and it hands you a perfectly formatted, highly confident answer. It feels so good you just copy, paste, and ship it. But who takes the blame when that strategy fails? The AI doesn’t. You do.
Look at what’s happening in the corporate world. UBS just made AI proficiency a hard requirement for junior investment banking jobs. They don’t care about your degrees or certificates anymore. If you can’t prove you can actually use AI to get work done, you won’t even get an interview. Colleges are rethinking what they teach, and companies are rethinking who they hire. The market is drawing a line in the sand.
But here is the blind spot everyone misses: They aren’t just testing if you can ‘use’ AI. They are testing if you can command it and own the results. We don’t need more people who can operate software. We need people who are willing to look at an AI-generated answer and say, ‘This is wrong. This prompt is garbage. I am making the call, and I own the consequences.’
The ultimate human advantage isn’t knowing how to control AI. It’s having the guts to be accountable for the decisions AI can’t make.
The president at Fudan broke down the word AI in a way that stuck with me. First, AI is Awaken + I. The awakened self. Second, AI is Action + I. The acting self. It’s not about the artificial intelligence; it’s about the human intelligence operating it.
Anxiety won’t save you. Waiting won’t save you. The middle ground of half-learning is a trap no one is going to pull you out of. I started a new habit. Every morning, I wake up 30 minutes early. I don’t scroll news feeds. I don’t read tutorials. I just use AI to do one real piece of work. It might be ugly. It might be wrong. But I force myself to question it, judge it, and take responsibility for it.
Answers aren’t scarce anymore. What’s scarce is a brain that can ask a good question, the guts to make a decision, and the shoulders to bear the weight of that decision.
AI can do your work, but it can never be you. Let it do the heavy lifting. You be the human.
FAQ
Q: Isn't AI just a tool that reflects the user's intelligence?
A: It is, but it's also a sycophant. If you don't actively question its output, it will flatter your existing biases and create an illusion of mastery while your critical thinking atrophies.
Q: What's the practical implication for professionals right now?
A: Stop passively copy-pasting AI outputs. Treat AI like an intern: give it clear instructions, rigorously check its work for errors, and take full ownership of the final decision.
Q: What's the contrarian take?
A: The real danger of AI isn't job displacement—it's cognitive offloading. We are trading our critical thinking and accountability for the cheap comfort of a confident, machine-generated answer.