Clean Code, Clear Thinking: Expanding Craftsmanship in the AI Era

Back in the 2010s, I wrote an article stating that software craftsmanship was the need of the hour.

Back then, the debate was whether craftsmanship slowed down delivery — whether “elegant code” was self-indulgent artistry that came at the cost of speed.

My view was simple:

• Craftsmanship is about customer value, not just elegant code.

• Well-crafted software respects end-users and future maintainers alike.

• Culture, not just practices, sustains innovation.

That thinking still holds true. But in the AI era, the canvas of craftsmanship has expanded.

Craftsmanship, Then (2010s)

• Keeping code clean, lean, scalable, and maintainable.

• Balancing delivery speed with structural quality.

• Building a culture where devs cared about why they were coding, not just what.

It wasn’t just about clean code — it was about clear thinking behind the code.

Craftsmanship, Now (2025)

AI has shifted the landscape. Tools reduce cycle times in coding, testing, integration, and deployment. But speed of generation ≠ speed of delivery.

Tech debt, bugs, and misunderstood requirements will still surface — maybe in new ways.

So the challenges — quality, trust, clarity — still remain. What’s new is the context: we now uphold them in an AI-augmented workflow.

Craftsmanship in the AI Age Means:

• Curating AI input — knowing how to prompt AI, and what system context to give it, so it produces clean, lean, scalable, and maintainable code.

• Curating AI output — knowing when to trust it, when to question it, when to override it.

• Keeping systems human-understandable — hybrid human+AI codebases must stay debuggable, extendable, maintainable.

• Extending beyond application code — into data pipelines, deployment scripts, test automation.

• Elevating collaboration — not just pair programming and code reviews → but pairing with AI and prompt reviews.

• Balancing speed with responsibility — embedding governance, compliance, ethics into prompts & pipelines to deliver secure, trustworthy, bias-aware software.

A Shift in Responsibility

Craftsmanship today is still about clean code — but it’s also about expanding clear thinking into new areas.

If the 2010s debate was working software vs. well-crafted software...
Then the 2020s debate is AI-generated output vs. human-guided craftsmanship.

Both rest on the same foundation: utility, quality, and respect. Respect for the customer, for peers who inherit systems, and now — for the responsibility of guiding machines.

Over to you:

• What does well-crafted software mean to you today?

• How do you practice craftsmanship in your AI-augmented workflows?