Addy Osmani on the next two years of software engineering

I recently came across an article that I think every developer in our community needs to read, whether you're just starting out or you've been building for years.

If you've been in web development for any length of time, you've likely come across Addy Osmani's work. He spent nearly 14 years at Google, leading developer experiences in Chrome, where he worked on DevTools, Lighthouse, and Core Web Vitals, before moving to Google Cloud AI as a director. He recently left Google after sharing 21 lessons from his time there, a piece worth reading in its own right. Beyond his work at Google, he's authored influential resources like Learning JavaScript Design Patterns and Leading Effective Engineering Teams, and has been a prominent voice in web performance and developer tooling for over a decade. When Addy writes about where the industry is heading, it's worth paying attention.

His latest piece, The Next Two Years of Software Engineering, examines five critical questions facing our industry: the future of junior developer hiring, the evolution of essential skills, how developer roles are changing, the specialist vs. generalist debate, and whether traditional education pathways still make sense.

Why this matters for the Frontend Mentor community

The themes Addy explores sit at the heart of what we're building here. A few points stood out to me:

On the junior developer question

Addy highlights research showing junior developer hiring has dropped significantly as AI tools become more prevalent. But he also presents the counter-scenario, that AI could unlock demand for developers across every industry. His advice for junior developers? Make yourself AI-proficient and versatile. Demonstrate that you can match a small team's output, but understand and explain every line of code. This is exactly why I believe building real projects matters more than ever. You need proof that you can deliver, not just follow tutorials. We recently wrote about how to get a programming job in 2026, and many of Addy's points echo what we're seeing.

On skills and fundamentals

There's a tension Addy identifies that I think about constantly. If AI writes most of your code, do you lose the ability to evaluate whether it's actually good? His suggestion is to use AI as a learning tool, not a crutch. Implement projects twice, once with AI assistance and once without. This approach aligns with how we've designed our challenges. The goal isn't just to ship something. It's to genuinely understand what you're building. If you're figuring out how to balance AI tools with actually learning, we've written a guide on using AI coding assistants as a beginner that covers this in more detail.

On the value of a portfolio

Throughout the article, Addy returns to a consistent theme: demonstrated skills matter more than credentials. Build a compelling portfolio with substantial projects and good documentation. Be active in the developer community. This has always been my belief. A well-built project that you can talk about intelligently will open more doors than a certificate ever will.

The bottom line

The industry is changing fast. AI isn't going away, and the developers who thrive will be those who use it to do more while still deeply understanding their craft. Building real projects, understanding fundamentals, and continuously learning have always mattered. They just matter more now.

I'd encourage everyone in our community to read Addy's full article. It's one of the clearest assessments of where we're heading that I've seen.

Read the full article: The Next Two Years of Software Engineering →