The software development community has split into two distinct camps, each with radically different approaches to AI tools.
On one side are the AI-first developers who have fully embraced tools like Cursor and V0. They ship code rapidly using AI assistance and believe this represents the inevitable future of programming. For them, manually writing code or searching Stack Overflow feels old—everything should be one click away.
The other camp, often consisting of more experienced developers, remains deliberately "anti-AI." They refuse to give control to ChatGPT, viewing it as a threat to the craft of programming. While some of this resistance stems from discomfort with rapid technological change, their concerns about skill atrophy aren't entirely unfounded. This divide creates a dangerous blind spot: What happens to new developers caught in the middle?
How should they learn how to code? Or how should they even code?
After considerable reflection, I believe the answer lies not in choosing sides, but in strategic balance. Here's the approach I'm taking:
My goal is to learn/adapt both, simultaneously. Learning about database fundamentals, networking, operating systems etc is fundamental so I would continue to dedicate some time every week to study and build small systems WITHOUT the aid of AI. The point here would be to struggle, deliberately.
At the same time, if I get any new idea, I would go ahead and build it in cursor. Here speed is a priority and getting something out there is more important than learning.
The future belongs to developers who can think deeply about systems and move quickly with modern tools. By consciously developing both skill sets, new developers can future-proof their careers while staying relevant in today's fast-moving landscape.