I watched a developer admit AI killed his business model. Here's why I'm not worried.

Last week, Jeffrey Way — the founder of Laracasts and one of the most respected voices in the Laravel community — posted a video that hit different.

He announced 40% layoffs at Laracasts. In the same breath, he admitted he's having more fun programming than ever before.

That paradox stuck with me. Because I've been living it too.

The business reality vs the personal reality

Jeffrey's business model is teaching developers how to write code. AI's business model is writing code for you. The collision was inevitable.

Stack Overflow's traffic is back to first-month-of-existence levels. Who waits 24 hours for a stranger to maybe answer your question when Claude gives you a working solution in 30 seconds?

But here's the weird part: Jeffrey says he's never enjoyed coding more. No more end-of-day headaches. No more Stripe webhook nightmares eating entire afternoons. The tedious stuff that used to drain him? Gone.

I feel the same way.

My "oh shit" moment

For me, the shift happened with Claude Code and Opus. Not gradually — suddenly. One day I was debugging a Laravel migration issue the old way (Stack Overflow, docs, trial and error). The next day I was watching an AI fix it in seconds while explaining exactly why it broke.

Then came agentic coding tools. Now I have an AI assistant that doesn't just answer questions — it executes. It writes the migration, runs the tests, commits the code.

The gap between "idea" and "working feature" collapsed.

The knowledge question

Jeffrey mentioned something that's been nagging at me too: was all that learning for nothing?

I spent years going from .NET to Laravel. Learning Eloquent patterns, mastering Blade templates, understanding the request lifecycle. And now an AI can write better Laravel code than I could after years of practice.

So why would anyone deeply learn React today? Or any programming language, really?

Here's where I land: knowing the fundamentals still matters — for now. When I review AI-generated code, my experience tells me when something's overcomplicated or when there's a simpler pattern. Jeffrey spends the rest of his day refactoring what the AI produces in 20 minutes. That refactoring requires human judgment built on years of experience.

But I'm honest with myself: there's probably a point where that stops mattering. We're not there yet. But it's coming.

My workflow: before and after

Before AI:

Building a feature took weeks. Write the specs. Design the database. Create migrations. Build the controllers, the views, the tests. Hit a wall with some library you've never used. Spend two days on documentation. Realize you architected it wrong. Refactor. Repeat.

After AI:

The most valuable work is now upfront: defining what you actually want. Clear specs. Good prompts. Understanding your own requirements deeply enough to communicate them.

The building part? That's the fast part now. The debugging? AI handles it. The "how does this library work?" research? AI explains it while implementing it.

A feature that would've taken me two weeks now takes a day. Sometimes less.

Do I review every line?

Jeffrey says he reviews every file, every line. His gut tells him the junk will compound if he doesn't.

I'm less strict — and I'll admit that might bite me later.

For production apps, yes: I still do proper PRs, run tests, review the important parts. But for side projects? Vibe coding experiments? I trust it more than I probably should.

The honest truth: I'm more likely to assume it's correct and move on. The AI hasn't burned me badly enough yet to change that habit.

What I don't miss

Here's what surprised me: I don't miss "real" programming.

I thought I would. I got into this because I liked solving puzzles, same as Jeffrey. The satisfaction of finally cracking a bug after hours of digging.

But you know what I like more? Seeing my ideas come to life.

The puzzle-solving was never the goal — it was the tax I paid to build things. Now that tax is lower. I can focus on the creative part: what should this thing do? How should it feel? What problem does it actually solve?

I'm building more, shipping faster, and enjoying it more. The identity crisis Jeffrey described in mid-2025? I went through it too. But I came out the other side.

The junior developer question

What about people just starting out?

Part of me thinks: what a time to begin. So many barriers removed. You can build real things from day one instead of spending months on fundamentals before you ship anything.

But I'm genuinely curious how they see it. Maybe they'll trust AI output 100% and focus on the ethics of what gets built rather than how it's built. Maybe they'll go deep on the math behind LLMs because that's where the real leverage is now.

Either way, the path into this industry looks nothing like the one I took. Whether that's better or worse, I honestly don't know.

The only way through

Jeffrey ended his video with a question: how are you feeling about this right now?

Here's my answer: I leaned in.

Not because I've given up. Because I've accepted reality. The tools are here. They're getting better fast. Pretending otherwise doesn't protect my skills — it just makes me slower.

The developers who adapt will build circles around the ones who don't. I've seen it already. I've felt it in my own productivity.

Is there something lost? Probably. The craft of hand-written code. The deep satisfaction of solving something purely through your own understanding.

But there's also something gained: more time for the parts that matter. Better ideas shipped faster. Less time fighting the machine, more time directing it.

Jeffrey's right. This isn't a trend. This is just how it's done now.

The question isn't whether to get on board. It's how fast.

What's your experience with AI coding tools? Are you all-in, skeptical, or somewhere in between? I'd love to hear how your workflow has changed.

I offer hands-on consulting to help you resolve technical challenges and improve your CMS implementations.

Get in touch if you'd like support diagnosing or upgrading your setup with confidence.

Learn more