What Vibe Coding Taught Me About AI, Product Design, and User Trust

Lately, I have been using Visual Studio Code more than I ever expected.

As a Product Manager, VS Code has not always been one of my go-to tools. But recently, I wanted to sharpen my AI skills, build something personal, and better understand how AI is changing the way people create. That curiosity led me into something people are calling “vibe coding.”

For me, vibe coding means using AI to help turn an idea into something real. You may not know exactly how to code everything yourself, but you know what you want to build, you can describe it clearly, and you are willing to keep experimenting until the vision comes to life.

Why I Decided to Build My Own Website

It started with a simple goal: I wanted to create a personal brand website.

I could have used a common website builder like Wix or a pre-made template, but I wanted something different. I wanted the process to stretch me. I wanted to sharpen my technical understanding, practice using AI in a real project, and create a design that actually felt like me.

I did not want a website that looked like everyone else’s.

So instead of starting with a template, I started with a prompt.

Starting With ChatGPT

I went to my good friend ChatGPT and described what I wanted.

I shared my goals, the type of brand I wanted to build, and the feeling I wanted the website to have. Then I asked ChatGPT to help me come up with a design that matched my vision.

The first version was not perfect, but it was a strong start. After a few iterations, the design started getting closer and closer to what I had imagined. Each prompt helped me clarify my ideas. Each version made me more excited.

Eventually, I reached a point where I loved the design.

Then I had a new question: how do I actually build this?

Moving From Idea to Code

At first, I did not know where to start.

I asked ChatGPT how I could launch the website. It suggested a few options, including using WordPress and manually putting the website together with different sections and components.

I had used WordPress years ago, so I knew that approach could work. But honestly, it sounded like too much manual work.

With how much AI has evolved, I thought, there has to be an easier way.

So I asked ChatGPT a simple question:

“Can you code this for me?”

And it did.

I was amazed.

ChatGPT helped turn the design into code and packaged it into a folder that I could open in VS Code. It also walked me through the process step by step, from downloading VS Code to opening the project and understanding how to make changes.

That was the moment I started to feel like I was not just imagining a website anymore.

I was building one.

Discovering GitHub Copilot

Once I started working in VS Code, I discovered GitHub Copilot.

At first, I was excited. I thought, “Wow, let’s see what this can do.”

Copilot helped me make changes, push my project to GitHub, and keep improving the website. It made the technical process feel less intimidating. I was able to move faster, ask questions, fix issues, and keep building even though I am not an engineer.

That is what made the experience powerful.

AI was helping me close the gap between what I wanted to create and what I knew how to build.

Then I Hit a Wall

Just when I was gaining momentum, something happened.

I ran out of quota for the next 28 days.

At first, I was completely confused. I did not fully understand what quota meant, how I had used it up, or why I suddenly could not continue using the tool the way I had been using it.

I was right in the middle of building my website. I was on a roll. These AI tools had become my support system for piecing things together.

Then suddenly, I felt stuck.

I started investigating and found guidance that suggested I could purchase more quota. I followed the link, which took me to GitHub. I saw where additional usage could potentially be purchased, but then I hit another blocker: that option was not available to me at the time.

I was honestly disappointed.

The Product Experience Gap

Looking back, the biggest issue was not just that I ran out of quota.

The issue was that I did not fully understand the limit before I reached it.

If I had known there were usage limits, I would have used the tool more carefully. Some of the things I used Copilot for could have been avoided. I could have saved it for the moments where I really needed technical help.

That experience made me think about AI products differently.

Powerful features are not enough. Users also need clarity.

If a product has usage limits, the experience should make those limits easy to understand. It should show the user how much quota they have used, how much they have left, and what happens when they run out.

It should also give a clear warning before the user reaches the limit, not after.

I may have seen a message about usage at some point, but I did not fully understand what it meant. So I avoided it and kept working. That is a product opportunity.

What Could Have Made the Experience Better

Once a user runs out of quota, the product should not make them feel stranded.

Instead, it could guide them.

For example, it could use a agent to ask:

“Do you still need help finishing your project?”

Then, based on the user’s answer, it could recommend another path forward. It could suggest using Microsoft Copilot, ChatGPT, documentation, or a step-by-step guide to continue the work without waiting for the quota to renew.

That small moment of guidance could make a big difference.

It would tell the user, “We know you are blocked, but here is how you can keep going.”

That builds trust.

The Emotional Side of Product Design

This experience reminded me that product design is not just about features. It is also about how the product makes people feel.

Copilot helped me build confidence. It made me feel capable of doing something that once felt too technical. But when I ran out of quota and did not have a clear path forward, that confidence took a hit.

I eventually figured out how to finish my project without Copilot, but the experience changed how I felt about returning to it.

That matters.

When a product supports users during the exciting parts but leaves them stranded during the hard parts, it creates frustration. But when a product guides users through blockers, it creates loyalty.

The Product Lesson I Took Away

Vibe coding taught me more than how to build a website with AI.

It taught me that AI tools have the power to help people move from idea to execution faster than ever before. They can help non-engineers build, experiment, and learn technical skills in a more approachable way.

But the best AI products will not just help users create.

They will help users continue.

They will explain limits clearly. They will guide users when they are blocked. They will offer fallback paths when something does not work. And they will make users feel supported throughout the entire journey.

That is the real product lesson I took away from this experience.

AI can help people build faster.

But thoughtful product design helps people trust the journey.