Beyond Syntax and Speed: Why Critical Thinking Is Your Best Leverage in the AI Era

Early in my career, building software involved spending more time working through systems, writing logic, and progressing step by step. Over time, we drifted towards optimizing for syntax, semantics and speed. In service based orgs, interviews, learning paths, and day-to-day work became centered around languages, frameworks, and keeping up with what’s new across ecosystems like Angular and JavaScript.

Somewhere along the way, critical thinking started taking a back seat.

Now, with tools like Claude, Codex, Copilot and Cursor, the experience feels very different. It’s almost like having an experienced Google or Meta developer working alongside you, almost like a chief of staff.

And what these tools are doing is bringing fundamentals back into fashion.

Over some time now, I’ve been exploring where each of them really fits.

➡️ Claude works well when you needto understand the architecture of legacy code, migration\refactoring with a structured thinking, breaking down problems, reasoning through approaches before you start building.

➡️ Codex, just 6 months old, still evolving as a coding agent, more direct support in writing and executing code.

➡️ Copilot VS Editor, continues to be strong inside the development flow with its coding agent capabilities.

➡️ Cursor goes a step further, building on top of this with its own editor and cloud-backed workflows, making the experience more integrated end-to-end.

Different strengths, but the same underlying pattern of a strong coding agent. The leverage is real, but the outcome still depends on how clearly the talent using it thinks, how well they define the problem, and how effectively they are communicating. Which is why, in a slightly unexpected way, strong fundamentals are back in fashion.

This is also where leadership needs to respond differently. Teams need to be encouraged to explore, experiment, and build real context around these tools, not just use them at the surface level. At the same time, we need to rethink how talent is evaluated and developed, how we hire, how we train, and what we recognize as quality output.

When the cost of building and iterating comes down, expectations around speed, depth, and market responsiveness will naturally shift.

And the organisations that adapt their ways of working to this reality will move very differently from those that don’t.