How to Learn in the Age of AI
I built Ferryman to solve a simple problem: manual cross-posting is low-leverage work that drains your time. With 18,600+ posts ferried for 460+ users, the system is proven to help you grow on all your socials with zero extra effort. Start scaling your reach automatically. Use code “BREAKSTUFF” for 30 days free of the Creator plan.
We’re currently living through the most radical shift in the history of software development. Code generation models have become so frictionless that you can scaffold entire features, debug complex stack traces, and deploy apps without ever truly understanding the underlying mechanics. Because of this, the engineering community is splitting into two distinct camps. The first camp is using AI to automate away their critical thinking. The second camp is using AI to accelerate it.
If you want to thrive in this new landscape, you have to fundamentally shift how you interact with these models. Here are the three rules for using AI to accelerate your learning, not outsource it.
1. Think More, Not Less
It’s never been easier to think less, but it’s also never been more important to think more. The temptation with AI is to treat it as an escape hatch from cognitive heavy lifting throwing a prompt at a model, wait for its output, and move on. That is a trap that will slowly but surely degrade your skills. Instead of outsourcing your learning to the tool, you should be leaning on it to deepen your understanding.
Use it as a world-class, hyper-patient tutor. When it gives you a solution, hopefully it’s one you’ve architectured and crafted together. Ask it to explain the trade-offs, the edge cases, and the performance implications. The goal isn’t to let the AI do the thinking for you; it’s to use the AI to pull your own understanding to a higher level, faster than any textbook ever could.
The biggest differencec with AI is that you no longer need to spend the time finding the resources that has the answer you need; now you can just ask for that information, regardless of what it relates to, in one centralized place.
2. Tools Shouldn’t Be Your Ceiling
If you aren’t better than the tool, you are only as good as the tool. If your value as a developer is limited to writing boilerplate code, debugging simple syntax errors, or doing basic API integrations, you are effectively operating at the baseline of the current best model. If you’re operating at the level of the current model you’re only as good as the models that are released.
AI should enhance you as an engineer; it shouldn’t replace your technical judgment. You need to stay ahead of the models by using them to handle the low-level, repetitive tasks so you can free up your brain to focus on the high-level system design, product strategy, and complex problem-solving. If you don’t use the tool to climb higher, you are letting it become your ceiling.
3. Methods Change Your Goals Shouldn’t
At the end of the day, you have to remember that AI is simply a tool. If you look at the history of computer science, programming has undergone massive shifts every few decades. We moved from physical punch cards to assembly, from assembly to low-level languages, and from there to high-level abstractions. Every single leap was met with anxiety that some level of “real engineering” was dying. This is just the next stage of that evolutionary loop.
Look at the glass as half full, not half empty. The tool isn’t replacing you; it is freeing you up to do far more interesting, creative, and high-leverage work. It is saving you time, eliminating the tedious boilerplate, and allowing you to move toward your ultimate goals at an unprecedented speed. Writing raw code has always been just one small part of engineering. The true craft has always been, and will always be, problem-solving: the ability to come up with simple, novel solutions to complicated problems.
Where you can find me online:
Drop a like ❤️ and comment below if you made it to the end of the article.