AI can scaffold an app in seconds, but can it refactor that thousand-line React file when the first bug hits production?

In this episode, I sit down with Brian Jenney software engineer and program owner of the coding bootcamp Parsity, to draw a hard line between “code that runs” and “code that lasts.” From mentoring career-switchers to stress-testing AI in real-world pipelines, Brian shares why craftsmanship and product judgment still beat copy-paste prompts.

🔗 Essential Links (Start Here!):

Full Episodes: https://product-driven.captivate.fm/episodes

Connect with Matt Watson on LinkedIn: https://www.linkedin.com/in/mattwatsonkc/

Subscribe to the Newsletter: https://newsletter.productdriven.com/

Powered by Full Scale: https://fullscale.io/

GET THE PRODUCT DRIVEN BOOK: https://productdriven.com/book

Connect with Brian on LinkedIn :https://www.linkedin.com/in/brianjenney/

Check out Parsity here: https://www.parsity.io/

Key Discussion Points:

“You have to be smarter than the AI.” Why blindly shipping generated code is the fastest way to paint yourself into a technical-debt corner

The hidden risk of non-deterministic models—like failing a simple “greater-than” check in production tests

A training rule of thumb: no AI for your first months of study, then use it only to reinforce fundamentals—not replace them

The “house-of-cards” analogy for code quality and why maintainability still matters when AI writes the first draft

How Parsity’s tight-knit model turns plumbers, teachers, and even doctors into disciplined, product-minded engineers