AI Assisted Coding: How Spending 4x More on Code Quality Doubled Development Speed With Eduardo Ferro

AI Assisted Coding: How Spending 4x More on Code Quality Doubled Development Speed

What happens when you combine nearly 30 years of engineering experience with AI-assisted coding? In this episode, Eduardo Ferro shares his experiments showing that AI doesn't replace good practices—it amplifies them. The result: doubled productivity while spending four times more on code quality.

Vibe Coding vs Production-Grade AI Development

"Vibe coding is flow-driven, curiosity-based way of building software with AI. It's less about meticulously reviewing each line of code, and more about letting the AI steer the process—perfect for quick experiments, side projects, MVPs, and prototypes."

Edu draws a clear distinction between vibe coding and production AI development. Vibe coding is exploration-focused, where you let AI drive while you learn and discover. Production AI coding is goal-focused, with careful planning, spec definition, and identification of edge cases before implementation. Both use small, safe steps and continuous conversation with the AI, but production code demands architectural thinking, security analysis, and sustainability practices. The key insight is that even vibe coding benefits from engineering discipline—as experiments grow, you need sustainable practices to maintain flexibility.

How AI Doubled My Productivity

"I was investing four times more in refactoring, cleanup, deleting code, introducing new tests, improving testability, and security analysis than in generating new features. And at the same time, globally, I think I more or less doubled my pace of work."

Edu's two-month experiment with production code revealed a counterintuitive finding: by spending 4x more time on code quality activities—refactoring, cleanup, test improvement, and security analysis—he actually doubled his overall delivery speed. The secret lies in fast feedback loops. With AI, you can implement a feature, run automated code review, analyze security, prioritize improvements, and iterate—all within an hour. What used to be a day's work happens in a single focused session, and the quality improvements compound over time.

The Positive Spiral of Code Removal

"We removed code, so we removed all the features that were not being used. And whenever I remove this code, the next step is to automatically try to see, okay, can I simplify the architecture."

One of the most powerful practices Edu discovered is using AI to accelerate code removal. By connecting product analytics to identify unused features, then using AI to quickly remove them, you trigger a positive spiral: removing code makes architecture changes easier, easier architecture changes enable faster feature development, which leads to more opportunities for simplification. This creates a self-reinforcing cycle that humans historically have been reluctant to pursue because removal was as expensive as creation.

Preparing the System Before Introducing Change

"What I want to generate is this new functionality—how should I change my system to make it super easy to introduce this one? It's not about making the change, it's about making the change easy."

Edu describes a practice that was previously too expensive: preparing the system before introducing changes. By analyzing architecture decision records, understanding the existing design, and adapting the codebase first, new features become trivial to implement. AI makes this preparation cheap enough to do routinely. The result is systems that evolve cleanly rather than accumulating technical debt with each new feature.

AI as an Amplifier: The Double-Edged Sword

"AI is an amplifier. People who already know how to develop software well will continue to develop it well and faster. People who did not know how to develop software well will probably get in trouble much faster than they would otherwise."

Edu's central metaphor is AI as an amplifier—it doesn't replace engineering judgment, it magnifies its presence or absence. Teams with strong practices will see accelerated improvement; teams without them will generate technical debt faster than ever. This has implications beyond individual productivity: the market will be saturated with solutions, making product discovery and distribution channels more important than implementation capability.

In this episode, we refer to Edu's blog post Fast Feedback, Fast Features: My AI Assisted Coding Experiment and Vibe Coding by Gene Kim.

About Eduardo Ferro

Edu Ferro is Head of Engineering and Data Platform at ClarityAI, with nearly 30 years' experience. He helps teams deliver value through Lean, XP, and DevOps, blending technical depth with product thinking. Recently he explores AI-assisted product development, sharing insights and experiments on his site eferro.net.

You can connect with Edu Ferro on LinkedIn.