AI can speed up learning a lot. But if you let it do the thinking, you lose your problem-solving skills. Here’s the system I use for university + coding + projects so AI stays just a supporter.
3 basic rules so AI doesn’t make you lazy
Attempt first. Try to solve it for yourself first, even if you have no idea it forces you to think about it for some time.
Ask for clarity, not conclusions. I use AI for explanations and options, but I still decide and do the work.
Verify the results with what you would expect. Don’t take them for certain.
3 Workflows I actually use + prompts
1. “Explain it precisely + give me intuition”.
When something feels new (math, ML, econ) I want a clear mental model.
Prompt: “Explain [topic] precisely. Give:
the core idea in 3 sentences
the intuition
one simple numeric example
common misconceptions.”
2. Turn notes into practice
Passive reading feels productive but doesn’t really teach you the same way practicing does.
Prompt: “Based on these notes, create 10 exam-style questions (conceptual + calculation). Put answers at the end.”
3. Debugging + coding faster
This is where AI is most useful and most dangerous (especially if you lose overview of your project and the code AI produced for you).
Prompt: “Here’s my code + error. Don’t rewrite everything. Identify the most likely cause, propose the smallest fix, and explain why it works.”
The Takeaway
AI should speed up the boring parts: searching, rephrasing, generating drills, and debugging. But it shouldn’t replace your critical thinking and problem-solving. The learning still happens when you try, get feedback, and iterate.
💡 My Recommendation of the week
Two Videos I really liked recently:
Have a great week,
Chris