Launch an AI Side Hustle with Vibe Coding and a Problem-Solving Mindset

Have you ever experienced this? When a browser extension you rely on suddenly disappears, it feels as frustrating as when Ctrl+C stops working while writing a report. I recently faced this issue—I needed to open multiple links at once, but discovered my batch selection extension was gone.

Instead of waiting for a replacement, I used Claude and Cursor to create my own solution with just one prompt. Using this case as an example, this article will introduce how beginners can start their independent developer side hustle with AI.

The Problem & Solution

The functionality is simple: allow users to press a key while dragging their mouse to select multiple links. It's a Chrome browser extension that seems minor but is incredibly useful for daily web browsing.

Let me walk you through my conversation with Claude. First, I described my need for a feature that would allow selecting all links within a mouse-dragged selection box. Later, I realized my description was incomplete and clarified that the functionality should activate when pressing the Z key while dragging.

Since I have a Claude membership and believe the official web interface produces better results, I generated the code on Claude first, then transferred it to Cursor. In Cursor, you need to create a folder for each project, and Cursor helps break down Claude's generated code into different local files. The advantage of Cursor is that it synchronizes with your local environment, and as an AI agent, it requires you to accept its operations before saving. After preparing the folder, I opened the Chrome extension store, enabled developer mode, and imported the folder. When the import failed, I simply copied the error message and asked Cursor to fix it.

The final result included not only what I requested but two additional details the AI added: the mouse cursor changing to a crosshair when Z is pressed, and an initial floating window tip informing users they can press Z and drag to select multiple links. You see, even such a small requirement has many interaction details to learn from and adjust.

Lessons for Beginners

A common beginner's mistake is immediately getting caught up in questions like "who is my target user?" or "how can I get people to pay me?" This makes everything seem more complex, trying to tackle too many things at once, until you have no idea where to start.

I want to use this example to show everyone that the best starting point for finding needs is identifying what feels inconvenient in your own workflow. Focus on those small "friction points" in your daily life—whether it's a browser extension or some other tool—where you think "it would be great if it could work this way." These tiny points of friction are your best entry points.

This approach has several benefits:

First, small utilities like this are useful but often don't make money. When my extension disappeared, waiting for someone else to develop a replacement could take forever. Using AI to create your own solution is perfect for these scenarios.

Second, the functionality is small, making it ideal for beginners to practice. Even if you don't know development or code, AI can still help you complete it. You won't get stuck like when building a large app where you encounter many problems but don't know how to solve them. Starting with manageable projects teaches you the limitations of AI tools. From this small project, I learned a lot—for instance, I discovered that because websites have different frameworks, some with complex JavaScript styles, the extension might fail to recognize certain links. This involves QA testing in different web environments to identify edge cases. While generating code for standard search results was easy, covering edge cases required multiple revisions, and some issues remained unsolved. These are things you only learn by doing.

Third, everyone is busy with work and finding time for side projects is challenging. Starting with small optimizations ensures you're not wasting time, as you're genuinely improving your own workflow. Each time you press Z, drag your mouse, and open five or six links at once, you'll feel satisfied thinking: "Hey, I made this myself!" It gives you a sense of accomplishment, and you'll feel your time was well spent.

Developing a Problem-Solving Mindset

Beyond development skills, this approach cultivates your ability to identify needs. The more you focus on starting with small details and noticing inconveniences, the more you'll develop this observational skill in daily life. You might notice things like: "Why do I have to click three times to reach the page I want?" or "Why do I have to manually organize this data every time?" These small complaints are actually seeds of opportunity.

To be honest, there are many opportunities to earn modest income through independent development. Making significant money might require luck, skill, and insight, but smaller opportunities are abundant.

This is the approach I recommend for beginners: if you don't know how to develop or find needs, don't try to do everything at once. Start by optimizing small details in your own workflow, observe what's inconvenient, what could be better, then try using AI to create solutions while learning its limitations.

Finally, let me emphasize that the "one-person company" is a path many will eventually need to take. Whether you're being exploited at your company or slowly boiling like a frog in water, many seemingly stable jobs are no longer stable, and some jobs might only last three to five more years. Exploring how to create your own stability beyond company-provided cash flow is urgent. This isn't about selling anxiety, but rather recognizing that the frog in warming water can already hear it boiling—when are you planning to jump out of the pot?