When AI coding finally clicked

Somewhere in the middle of the chaos of this dark year, something shifted for me. AI went from feeling like an existential threat to my career to sparking the kind of excitement I hadn't felt in a long time.

The struggle was real

For the first half of 2025, I was in a real identity crisis. AI tools were exploding everywhere, disrupting everything at this blazing fast pace. People were excited. People were scared. People were mad. Confused.

I didn't know what to think. What to feel. I wasn't really sure what my role in all of this was. Or should be. Or would be.

I'd spent years building up expertise in web development and teaching what I'd learned to others. And then suddenly it felt like anyone could spin up a completely custom (and very poorly-built) website without knowing anything about code. And my company was working to be on the cutting edge of this revolution, to help these new creators build successful projects on the web.

What did that mean for me?

The communities I'd been part of for the last several years — the ones that had been so welcoming and inclusive to new developers — were splitting apart. Some folks were all-in on AI. Others were actively hostile to it. I found myself caught in the middle, feeling conflicted about where I stood.

Was I abandoning the inclusive environments that had welcomed me if I supported AI development? Was I on the wrong side of history?

For months, I struggled to answer these questions. And yet, to do my job well, I needed to figure this out — to embrace the change and help lead the way, or get out.

The moment it clicked

And then I read something. I can't remember exactly where or what it was. But it provided this spark that began to reframe how I thought about AI and web development.

The insight revolved around inclusivity. And it was a look into the past that helped me see the present more clearly.

The front-end revolution

A decade ago, a new revolution was underway — one that would completely redefine the role of a front-end developer. Before that, front-end developers wrote HTML, CSS, and maybe a little JavaScript. There was much talk of if these folks were developers. "Developers" were "real programmers" writing with "real programming languages." It was exclusive.

As the tech evolved, so did our communities. Very quickly, the API economy burst open. If front-end developers learned a little more JavaScript, they could build complete, full-stack applications. Front-end developers became developers. And I found myself teaching JavaScript to people who'd never written code beyond HTML and CSS before. Together, we widened the definition of what it meant to be a developer.

(This was an intentional simplification of the 2015-2020 time period. We have other things to discuss here.)

As I thought back to this shift, I realized that we're going through the exact same thing again.

History repeating itself

What I started to see was that we, once again, have an inclusivity problem. We redefined what it meant to be a developer over the last 10 years. And now, a new set of tools has enabled a new group of people who want in — they want to self-serve building full-stack applications.

And there's much resistance to the change.

Addressing the concerns with AI

I need to pause for a minute to consider a bit of nuance. While there are similar patterns to the front-end revolution and API boom, there are very real concerns that go beyond fear of change that need to be addressed.

We have very real challenges in front of us in security, sustainability, and ethics to address. And we need to work together on these.

I don't want to open up the political worm can here. So I'll just say this: there are things in our societies that demand resistance. That do only harm. And then there are things that can be great, but need to be handled with care (and proper guardrails). AI fits the latter category.

The real opportunity

Once I made the connection to patterns of the "developer" title gatekeeping, everything changed for me.

I started thinking about all the educational material and teaching that happened between 2015 and 2020 (and beyond). So much of it was focused on basic patterns, and tips and tricks that developers could pick up and use. It meant helping people who were new to code or coming from a design background level up their skills.

That is exactly where we are again.

Many (maybe most) people out there using AI tools have no idea what their AI agent is building for them. They just see that it works. And they don't care how. They're having fun. And that's fine. Tooling needs to get better to help these folks wield their power safely.

But there's also this huge contingent of people who want to build with AI and want to build something serious and scalable with polish.

These are the people who need new learning paths. They need to understand when to trust the AI output and when to dig deeper. They need to learn patterns and best practices in a world where the tools used to build are fundamentally different.

What this means for web development

As soon as I saw this, I got excited again. Because this is exactly the kind of challenge I love — figuring out how to teach people new skills in a changing landscape.

I started seeing new opportunities everywhere:

• New learning paths: We need to figure out how to teach web development in the age of AI. Not "here's how to write a for loop from scratch," but "here's how to evaluate and improve AI-generated code."
• Different skill emphasis: Understanding problems deeply is becoming more valuable than memorizing syntax. Being able to architect solutions and spot issues matters more than typing code.
• Wider access: Just like 10 years ago, we're about to see a wave of new developers who come to code through a different path. They are excited and hungry to learn.
• Opportunities everywhere: For those of us who are willing to explore and learn in public, there's an endless number of workflows and tools and problems to dig into.

Moving forward

I'm not saying everyone needs to develop with AI tools. Just like not everyone needs to use React or vanilla CSS or indent with spaces or whatever other technology debate you want to pick. You choose the best solution for the unique situation.

Sometimes the simple tool is the best tool. I reach for a screwdriver just as often as a power drill. It depends on the job.

The right amount of skepticism is healthy and important. But that skepticism was holding me back from the opportunity. Instead, I began to find it more productive to use the skepticism to help developers find the best ways to use AI tools effectively.

The mission remains the same

There's so much to explore here. The landscape is evolving daily.

That's exciting (to me)!

I've spent years teaching web development and helping people level up their skills. This doesn't make that work irrelevant. It just means the curriculum needs to evolve. The patterns need to adapt. The focus needs to shift a bit.

At my core, my mission remains the same: to help developers build better websites.