Syntax - Tasty Web Development Treats · Wes Bos & Scott Tolinski - Full Stack JavaScript Web Developers

986: Does Code Quality Matter Anymore?

March 11, 2026·59 min·5 clips
Wes Bos warns that outsourcing the 'brain job' to AI risks making developers irrelevant in software development.
Wes Bos opens with a brief update on his ongoing search for a personal note-taking and knowledge management system, describing incremental progress with a second-brain approach. The first listener question comes from Alex, asking whether the availability of native HTML dialog and popover elements means developers are back to using display:none on hamburger menu buttons on smaller screens. Wes and Scott conclude this is essentially a non-issue, confirming that yes, display:none on the button when the nav is shown at full width is simply how it has always worked, and that the newer CSS techniques for animating popover and dialog elements — including starting-style and allow-discrete properties — represent the future but are not yet fully supported across browsers, keeping most developers reliant on JavaScript class toggling for backward compatibility. The discussion includes a detour into an observation Wes made about Vercel's iOS app using the Vibrate API via a creative workaround: toggling a hidden checkbox switch rapidly to trigger haptic feedback, since iOS does not expose the standard navigator.vibrate API that Chrome on Android supports. The main substantive question comes from Krasimir, asking whether code quality and organization still matter when AI is reading and writing most of the code. Wes and Scott engage seriously with this question, distinguishing between the idea that AI can tolerate messy code and the question of whether it should. Both argue that well-organized code still matters for human comprehension, debugging, and long-term maintainability, and that the claim that AI makes code quality irrelevant reflects a misunderstanding of what code organization is for. They acknowledge that some conventions may evolve as AI tooling changes, but resist the inference that developers can simply generate slop and rely on AI to navigate it. The episode also touches on Effect.ts, browser support strategies for new CSS features, and whether dedicated layout systems are still necessary given the maturity of native Flexbox and Grid.

As heard by us

A practical web-dev check-in on AI, layout, and browser-support tradeoffs.

This episode keeps its focus on practical tradeoffs: AI-era code organization, whether Flexbox and Grid still matter, and where newer HTML and CSS features collide with browser support limits.

Read the full review in PlayNext →

Why you'd press play

Wondering whether code organization still matters when AI writes more of the glue?

Read the full recommendation in PlayNext →
Listen to the show on