5 Things an Office Worker Did to Earn Their First 10,000 Yen with an AI Side Hustle

Two years ago, I was an office worker who could only use Excel and Word.

I couldn't program. I couldn't read English. I had zero experience with side hustles. Every time I heard about "making money with AI," I thought it was for people like engineers or former consultants.

But now, I earn about 70,000 to 100,000 yen a month in side income alongside my main job.

However, it wasn't smooth sailing from the start, and not everything is going perfectly even now. It took me three months to earn my first 10,000 yen. In this post, I will write about what I did during those three months.

1. I used the free version thoroughly for three weeks before paying for the premium version

I first tried ChatGPT in the spring of 2024. At the time, I was still using the free version, and I remained in a state of "it's amazing, but I don't know what to use it for."

For the first week or two, I only used it for things like "write a blog post" or "fix this email." Honestly, I didn't really understand what it could do.

What changed was when I started "asking it to do the specific tasks of my job directly." Organizing meeting minutes, creating drafts for reports, summarizing multiple notes into bullet points. There were surprisingly many tasks for an office worker that could be supplemented by AI.

After using the free version for three weeks and getting a feel for how it could be used for work, I switched to the paid plan for about 3,000 yen a month. Looking back, I'm glad I didn't rush into paying for it.

2. I narrowed my side hustle down to one type — I wasted a month trying to do everything at once

For the first month after deciding to start a side hustle, I didn't earn anything.

The reason was that I was trying to do everything. AI writing, image generation, transcription, prompt selling—I researched various things, tried them little by little, and ended up leaving them all half-finished.

Honestly, it was exhausting.

The turning point was deciding to "choose only one thing that combines skills I can use right now with AI." Tasks I do daily as an office worker—organizing text, summarizing documents, light proofreading. I thought if I did those with "AI + my own hands," it could become work that could be outsourced.

3. My first project was "audio transcription + editing"

When I looked on CrowdWorks, transcription-related jobs were relatively common.

The process was to do a rough transcription using AI (speech recognition tools), then read and refine it myself before delivery. The unit price per project was low. About 1,000 to 2,000 yen. Even so, I chose it as my first "experience of getting paid."

When I delivered my first project, I received a reply from the client saying, "You summarized it in an easy-to-understand way." More than the money, the feeling that "it functioned properly as a job" became the reason I kept going.

To digress a little, I feel like it was around this time that the sense of "AI is the draft, I am the finisher" became established. That influenced my later direction as well.

4. I had zero earnings for the first 43 days

It took 43 days from starting my side hustle until I received my first earnings. I wasn't counting exactly, but I remembered thinking "finally," so I went back through my planner to check.

The reason I didn't quit during that time is simple: I believed that "I can't make a judgment unless I keep going." Whether things weren't working out because of my own issues, the method, or the market itself—I wouldn't know unless I tried.

It's easy to find reasons to quit a side hustle. But I thought it was too early to make a judgment after just over a month.

5. Once I shifted my focus to "writing," I exceeded 10,000 yen per month for the first time in my third month

After experiencing a few transcription jobs, I started trying to "write articles myself and post them on note."

I used AI to create a draft, added my own experiences and thoughts, and then reread and polished it. This flow suited me well. Since I was used to the task of organizing documents as an office worker, it wasn't as difficult as I had thought.

In the third month of my side hustle, the total earnings I received that month exceeded 10,000 yen for the first time. This figure combined jobs from CrowdWorks and sales from note.

It's not a large amount. But it was the month I felt the realization that "it is working as a system."

The most important thing I want to convey with this story

Deciding "what I will do using AI" was more important than technical skills.

Knowing tools and earning money with tools are different things. Looking back, the first month I spent just studying tools, skipping the step of "deciding what I will do" that lies in between, wasn't even a preparation period—it was just procrastination.

To be honest about reproducibility, I don't know if the same flow will lead to the same results. "10,000 yen in 3 months" is my case, and it changes completely depending on the time you can use, what you are good at, and the type of work you choose. However, the idea that "you can't do an AI side hustle without IT skills" did not apply to me, at least.

After exceeding 10,000 yen a month, I hit the next wall. I felt that just "using tools" would hit a ceiling. I wrote about what I changed after that—shifting from the "user" side of AI to the "designer" side—in the next article.