"AI Side Hustle for Hospital Administration" - What Started Moving Beyond the Predictions [True Story Vol. 16]

A True Story of an AI Side Hustle by a Former Hospital Administrator — Part 1, Complete

It was a pitch-black room.

As usual, I was lying in bed, scrolling through my smartphone before going to sleep.

I was scrolling through articles about side hustles.

Some articles are interesting. Some are helpful. But as I read them, there are moments when something shines through.

In those moments, a feeling akin to resignation washes over me.

While scrolling, I suddenly thought.

Isn't my own writing being seen by someone in the same way?

Isn't it being perceived as not genuine?

I put my phone down.

I stared at the dark ceiling.

The question turned toward myself

We have entered an era where anyone can easily create something decent using AI.

If you refine your prompts, AI will generate plausible text. These days, such techniques are everywhere.

If a person is just posting generated text, aren't they unnecessary? Couldn't the AI just post directly to Note?

I was writing in such an era.

But I realized it once I started writing.

The moment I tried to convey things well, it became too polished. When I thought too much, it became too clean. When it became clean, it became uninteresting.

It's not interesting unless you leave some dirt behind. It doesn't look genuine unless it's a bit messy.

Writing that is too polished becomes indistinguishable from what AI writes.

The moment I tried to summarize know-how. The moment I looked for words to attract people. I was in the same place.

If you do it without thinking, the same thing comes out.

I've come to understand how to use AI. But the feeling of becoming the same as everyone else didn't go away.

Since I started writing myself, rather than just watching from the outside, that fear became even more real.

It was about what to lead

Vol. 6I wrote this in.

I am the one in the lead. AI only makes suggestions, and I am always the one who makes the decisions.

At that time, I thought that was the correct way to use it. In fact, it was.

But after writing 15 entries, I've started to see a deeper level.

I understood the 'leading' part. But it was about what to lead.

You can create images with AI. You can create videos. You can create music. Games, too, and apps.

I thought it had become convenient. But at the same time, I saw something else.

When I switched from handwriting to a pen tablet, the tools changed. But the difference in expression didn't disappear. In fact, the better the tool's performance, the more the difference in the human using it was exposed.

I thought AI was the same.

Every time tools change, the differences on the human side become visualized. The sense, the experience, what one wants to create.

That alone cannot be decided by a tool.

For example, think about a taxi. If you accurately convey the destination, time, and route, you won't go to the wrong place, you won't be late, and you won't take a detour. AI is like that. There's no need to overthink it. If you convey it accurately, it moves accurately.

But once I became able to convey things accurately, I noticed something.

A taxi only delivers what is loaded into it. No matter how accurately it reaches the destination, if what is loaded is empty, what is delivered is also empty.

What I was loading was different

The true nature of the fear began to appear.

I was trying to systematize it. The moment I did that, it became thin.

There are several scenes I remember.

When the hospital system went down, hundreds of people rushed to the reception desk. If you mess up the initial response, it turns into a riot-like panic. Fights between patients also break out. I actually experienced this when I was a rookie.

So I created a manual. And I learned something from creating it. Even if there is a manual, it won't function unless there is someone who can take command. You need to cultivate that person on a regular basis. You can't understand such things unless you are on the front lines. It's a story that exists outside the manual.

The same was true for guiding doctors on increasing revenue. At first, they gave me looks like, 'What does an administrator know?' Even so, I visited them many times. I repeated the explanations. Trust could only be built up over time.

There is also the story of a veteran part-time worker. There was a workplace where the manual existed only in that person's head. To protect their own place, they wouldn't teach anyone. Since the management side didn't grasp the operations either, the problems weren't visualized.

I turned it into a manual and generalized the work. Anyone could do it to some extent. But at the same time, 'It has to be that person' changed to 'It doesn't have to be that person anymore.'

That irony weighed on me now.

The more you become able to use AI, the same thing happens. The more tools become usable by anyone, the more 'it doesn't have to be that person' spreads. What remains is only what cannot be produced by tools.

Only the details of the experience were different

I once had AI predict my article.Vol. 15It's a story I wrote in. What came out was a plausible article. The writing style, the structure, the sense of temperature. But the content was something completely different.

What AI cannot predict is the details of the experience. The emotions of that moment. The sensations that only someone on the front lines knows.

When I wrote with that in mind, something was different. I, the writer, was shaken. Only then did I feel it became genuine.

At the same time, I intentionally avoided the predicted development. I tried to bring my experience and emotions to the forefront. But the more I polished it, the thinner it became. If I broke it down too much, it wouldn't be conveyed. That balance took the most time.

The moment I tried to make it look good, it became thin. That wasn't an AI problem. It was my problem.

I also fixed the place of delivery

Vol. 15I noticed one more thing in. The people I wanted to reach weren't on social media.

There are people who feel anxious about AI but can't ask anyone, so they search on Google. To reach those people, I changed the title. I changed the headings. I rebuilt the place of delivery.

Just doing the method correctly wasn't enough. I also had to make the place correct.

There was a place I could reach

In a world where anyone can use AI, the difference in tools becomes zero. When the difference becomes zero, what remains is only experience and the actual experience.

There is a paradox. The more equal tools become, the more only the difference in experience remains.

That is why the experience of hospital administration now has value. In the era before AI, that experience was completed only within the workplace. Only by combining it with AI did it become something that could be taken outside.

The things I have accumulated over 20 years at the hospital have finally met the conditions to become a weapon in the AI era. It cannot be imitated by people who do not have AI. It cannot be imitated by people who only have AI. There is a place that only people who have both experience and AI can reach.

What changed in 15 installments

I trusted the AI and took action. I made it admit its mistakes. But it was I who was wrong.

I started taking the lead. I learned the patterns. I went back and forth.

The location was off. The way I delivered it was off. I fixed it.

A response came back. The zero moved.

Noticing, fixing, and moving again. It was a repetition of that.

It wasn't just the numbers that changed.

Vol. 6I wrote this: "Use AI as a tool."

Vol. 15I wrote this: "Even if the tool changes, it is you who is using it."

And in Vol. 16, I saw one step further.

It is me who is using it. But there was something beyond that. It was about what I bring to the table when I use it.

Experience, emotion, warmth, and scenery. When I approached it with those, for the first time, I created something that only I could make.

The first part ends

What I gained in the first part was not profit.

I learned how to use AI. I learned how to deliver. But the biggest thing was that for the first time, I could believe that my own experience would become a weapon.

These 15 installments were a record of that. Days when things didn't go well, days when I stopped, days when I was off-track—I wrote it all down. Because it wasn't pretty, it was real.

There is a view that can only be seen after you have moved.

That night, I was looking at the dark ceiling. I didn't know if my articles were real. Even now, I still haven't found the answer. But I, who have moved, can see things I couldn't see back then. That much is certain.

There are people who have read these 15 installments so far. Stories of things not going well, stories of being off-track, and stories of continuing anyway. If you have read to the end, you have a reason to move, too. The tools are already in place.

All that's left is to approach it with your own experience.

Next time preview

In the second part, I will face that question.

In Vol. 17, I will write about that.

Honest current status

----------

Profit: 1,700 yen
CrowdWorks: Registration complete, waiting for projects
Coconala: Page improved, waiting for inquiries
X: 370 followers, currently used as a notification tool
note: 800 followers, main battlefield
Paid articles: Published and fully updated
Interaction with AI: Managed by dividing into processes
note series: Publication complete up to Vol. 16

----------

There was one thing AI could not predict.

It was me, having moved.

(Continues to Vol. 17.)

👉 Series articles are compiled in a magazine.

If you are thinking about a "side hustle with AI," let's try and fail together. Send questions to my X DM.