Is AI Killing the Joy of Building Websites?

I remember the first website I built. It was ugly. The layout broke in certain browsers. Half the images were missing. And I could not stop smiling.

I had made something that existed on the internet. I had typed things and the world could see them. Nothing felt more like magic than that.

Now, in 2026, an AI tool can build that same website in four minutes. Same layout. Probably better. Definitely faster. And I sit here wondering – why does that make me feel a little sad?

That question has been everywhere lately. Developer Reddit is full of it. Hacker News discussions run hundreds of comments deep. “AI has taken the fun out of programming.” “I do not know what my job even is anymore.” “Is this worth learning now?” The anxiety is real, and it is loud, and I do not think dismissing it serves anyone.

So I want to actually think through it. Not with the usual “AI is just a tool” response that feels true but empty. I want to sit with the uncomfortable parts.

---

The Speed vs. Soul Trade-Off

I have been building websites for over a decade. In that time, I have watched the tools get better, faster, and smarter at an almost frightening pace. Every year, something that used to take a week takes a day. Something that took a day takes an hour.

AI did not start this trend. But it has accelerated it to a point where the gap between “idea” and “shipped” is almost nothing now. And that should feel good. In many ways, it does. I ship faster. I iterate faster. I can say yes to clients I would have turned down before because the scope felt too large for the budget.

But I keep thinking about something my father used to say when I was a kid. He built furniture by hand. He had a workshop in the back of our house, and I would watch him spend entire weekends on a single chair. I once asked him why he did not just buy one. He said building it slowly was the point. By the time he finished, he knew every joint, every angle. He understood the whole thing.

There is a difference between assembling IKEA furniture and building a table with your hands. The result might look similar to someone sitting across the room. But what you carry away from the experience is completely different.

That is what I keep coming back to. Building things slowly taught me things that building things fast never could. I learned by making mistakes that cost me hours. I understood systems by being forced to fix what broke inside them. I developed instincts by hitting the same wall enough times that I learned to see it coming.

A young developer who has AI solve every problem from the start – do they build those same instincts? I genuinely do not know. Maybe they build different ones. Maybe better ones. But I notice I am asking the question more and more, and I do not have a clean answer yet.

When AI removes the productive struggle, does it also remove some of the learning that comes from that struggle?

---

When a Client Said “AI Can Do What You Do”

About eight months ago, a client said something to me that I have not been able to shake.

We were in a call, going over the scope of a project. It was a fairly involved build – custom functionality, some tricky integrations, a membership flow that needed careful thought. Good work. The kind I enjoy.

Halfway through, they stopped me and asked, “Can we not just use one of those AI tools? I read that they can build full apps now.”

My first reaction was fear. Not panic, but a cold, quiet kind of fear. The kind you feel when something you thought was permanent suddenly seems fragile.

My second reaction was curiosity. So I tried it. Right there, on the call, I showed them what those tools could actually produce for their use case. I wanted to know the real answer, not the comfortable one. (I have written about the specific AI tools I use in my own WordPress development work – that post goes deeper into what actually helps vs. what is mostly hype.)

The output was impressive. I will not pretend otherwise. For a simple version of what they needed, the AI got surprisingly far. Faster than I would have expected six months earlier. The code was clean. The structure made sense. A capable junior developer could have taken it somewhere real.

But here is what it missed: the client had a history. They had customers who had complained about a specific checkout behavior for two years. They had a team that operated a certain way. They had a brand voice that was specific and a little unusual. They had opinions about what their competitors were doing wrong that shaped how they wanted their own product to feel.

The AI built what they asked for. I was there to figure out what they actually needed.

AI built what they asked for. I was there to figure out what they actually needed.

This is not a theoretical distinction. There is a story circulating right now on Reddit – a company whose customer built an AI-powered replacement for their product. The customer felt proud of it. Left. Then came back six months later. The thing they built worked fine, technically. But it was missing everything that made the original worth using. The edges. The judgment calls. The quiet decisions someone made a hundred times before the customer ever opened the app.

Those things do not live in the code. They live in the people who built it. In the accumulated thinking that shaped every small decision.

That client of mine stayed. We built something real together. And at the end of the project, they told me the most valuable part of working with me was not what I produced but what I helped them figure out about what they actually wanted.

That is a thing AI cannot do yet. Not because it is not smart enough. But because it requires being in the room, asking the second and third follow-up question, noticing the hesitation, reading what is not being said.

---

The Vibe Coding Conversation

There is a new term floating around: vibe coding. The idea is that you describe what you want in natural language, and AI writes the code, and you just… keep going. You never look at the implementation. You trust the output and move forward.

Some people love this. And honestly, for certain things, I get it. Prototyping. Testing an idea fast. Moving through the boring scaffolding to get to the interesting parts.

But there is a version of vibe coding that worries me. The version where someone ships a product to real users without ever understanding how it works. Where the first time they encounter a real problem is when a customer hits it. Where they cannot debug what they cannot read.

This is not hypothetical. I have talked to founders who are deep in this situation. Their AI-generated product worked beautifully in demos. Then it hit an edge case in production that they had no idea how to diagnose, let alone fix. They had to go back to square one – not because the AI failed, but because they had no foundation to stand on when things got complicated.

Vibe coding shifts the maker movement in interesting ways. It lowers the floor for who can build something. That is genuinely exciting. But it also creates a generation of builders who may not be equipped for the hard problems that always show up eventually. I explored this angle in what skills actually matter when AI agents can handle the basics – the answer is not what most people expect.

I think the most powerful position is to understand both – to know enough to build with your hands when it matters, and to know when to reach for the faster tool. That combination is not easy to develop. But it is worth developing.

---

What AI Cannot Replace

I want to be honest here. I am not going to give you a list of things “only humans can do” and then pretend AI will never get there. Maybe it will. Maybe faster than any of us expects. I do not know.

But right now, in my actual work, these are the things that still require a person.

Listening. Real listening. Not processing words but understanding the feeling underneath them. When a client says “I want something clean,” that word means fifty different things. Part of my job is figuring out which one they mean, and that requires reading the room, asking the right follow-up, noticing what makes them lean forward in their chair and what makes them go quiet.

Taste. This one is harder to talk about without sounding arrogant, but it is real. Knowing when something looks right. Knowing when a design feels off even before you can articulate why. That sense is built from years of looking at things, making things, having opinions about them, changing those opinions. It is not something you can download.

The relationship itself. Some of my best clients have stayed for six, seven, eight years. Not because I am the cheapest option or the fastest. Because there is a trust built from working together – failing sometimes, solving problems together, knowing each other well enough that fewer words are needed each time. That continuity has value that no single deliverable can capture.

WordPress has survived every wave of “this will kill it.” Website builders. Drag-and-drop page builders. No-code tools. The various Webflow-shaped alternatives. It is still here, still powering a large portion of the internet, still growing. Not because it is perfect, but because it has something around it – a community, an ecosystem, a culture of people who genuinely care – that outlasts any single tool or trend.

I think the same is true of good developers. Not developers who are attached to specific technologies. But developers who are curious, who listen well, who care about what they are making and who it serves. That kind of person will always have work. Because technology changes. Human problems mostly do not.

---

My Honest Answer

I have thought about this question a lot over the last year. Is AI killing the joy of building websites?

My honest answer is: no. But it is changing where the joy comes from. And that shift is real enough that it makes sense to grieve the old version a little before embracing the new one.

The joy used to come from the act of making. From writing something and watching it work. From solving a problem that took me three days. From figuring something out through sheer effort when I had no other option. That joy was real, and I miss parts of it.

The joy now comes from somewhere slightly different. It comes from solving someone’s actual problem. From shipping something that makes a real difference to a real person’s work. From the conversation before the build, not just the build itself. From sitting with a client at the end of a project and seeing that what we made actually changed something for them.

AI helps me get to that part faster. And honestly? That part was always better. I just had to spend a lot of time on other things to get there.

I think the developers who feel like AI has killed their joy are the ones who only loved the typing. The ones who found meaning in the specific act of writing code, line by line, watching it compile and run. That part is genuinely being compressed. If that was everything to you, then yes, this must feel like loss. That is real.

But for those of us who loved the thinking – the figuring out, the listening, the shaping of something from nothing into something useful – there is more space for that now, not less. The parts I used to spend on boilerplate and setup, I now spend on the harder and more interesting questions.

That feels like a trade I can make.

---

The Craft Is Not in the Code

My father never cared about the specific tool he used to make a chair. He cared about the chair. He cared about the person who would sit in it for the next twenty years. He cared about whether it would actually last.

That is where I have landed after a year of thinking about this.

The craft is not in the code. The craft is in the caring – about what you are making, who it is for, and whether it will actually help them. The code is just the current best way to express that caring. When there is a better way, use it.

Use the tools that help you do that. All of them. Whatever works. Do not be romantic about the slow way when the fast way serves the person you are trying to help.

And if you find yourself grieving what used to take longer – I get it. Sit with that for a minute. It is a real feeling. It means something. It means you cared about the thing you were doing, and that caring does not disappear just because the process changes. That caring is actually the most valuable thing you bring. I wrote about a related idea in building quietly and letting the work speak – it is the same principle, applied differently.

Carry that caring forward into what you build next.

Then build something that matters.