The One Habit I Changed Before Using AI to Improve Efficiency

Hello, this is Reiji Kuga.

Since I started game development, there is something that surprised me a little. The difficult part wasn't Unity. It wasn't AI, and it wasn't Codex either.

It was starting where I left off yesterday.

"Wait, how far did I get yesterday?"

It takes more time than I thought just to remember that. For me, coming home after my day job and working on game development for about an hour at night, that time was more wasteful than I could have imagined.

"The time spent remembering" was the longest part

Right now, I am remaking the metaverse drama "Crossroad," which I produced for Project Alive, into a new form as a game. I've started using Unity and, with the help of AI, I am gradually moving forward with development.

However, if I take a few days off from working, my brain resets in a strange way. I start searching for my yesterday-self before I can even make the game, wondering, "Why did I decide on that specification?" or "Was this fix already finished?"

I used to think that was just part of the work. However, on days when I only have one hour to use at night, those first 15 or 20 minutes disappearing just for checking things is actually quite significant, I finally realized one day.

It wasn't that the development itself was slow; it was that it took too long to resume.

"Seeming convenient" was the most dangerous thing

In an attempt to create even a little more time, I once tried running Codex remotely from my smartphone. I thought it would be efficient if I could make progress even during my work breaks. It was that kind of lighthearted feeling.

But in reality, it was the opposite. I ended up spending more time checking things than actually making progress, wondering, "Which environment am I working in right now?" or "Is this change really reflected?"

I remember the moment I felt a chill, thinking, "I can definitely undo this, right?"

It wasn't that Codex was bad. It was that my way of using it wasn't good. Just because it's a convenient tool doesn't mean it's convenient in every situation.

Since that day, I decided to work on game development calmly only on my PC. Even if it looks like a detour, it's faster in the end. At least, that method suits me right now.

What I'm leaving behind isn't a daily report

Now, before I finish my work, I organize four things: what I did today, how far I've come, what I'll do first tomorrow, and the reason I stopped midway.

At first, I thought it was just a work log. Something to record the day's progress and feel a little relieved. That was about the extent of it.

But as I continued, it started to look a little different. This wasn't a note for looking back on today, but a note for handing over the continuation to my tomorrow-self.

The thing I least want to write about is the reason why I stopped midway. Days when I have to record that 'I barely made any progress today' are heavy on my heart. There are days when I struggle for hours, only to end up with more errors than I started with.

On days like that, I want to close my computer without writing anything at all.

It feels like I am recording my own incompetence. Yet, strangely enough, the thing that helps me most the next day is that very part I didn't want to write down.

'Ah, so this is where I got stuck yesterday.'

Just knowing that saves me from having to search for the same spot all over again. Even if it isn't solved, just having a record of where I got lost makes it easier to start the next session.

'Crossroad' wasn't always smooth sailing either

While working on game development, I suddenly remembered the production of 'Crossroad.' It took over three and a half years to complete. Along the way, there was a period where I stopped the release and rebuilt the work itself from scratch.

Looking back, it wasn't always smooth sailing every day. There were days when I only progressed a few pages in the script, and days when I ended up just opening the editing software. There were many nights when I felt like I hadn't accomplished anything at all.

Even so, the work never came to a complete standstill.

It wasn't because I was making huge strides every day. I think it was because I kept handing the continuation over to myself, saying, 'Let's start from here tomorrow.'

Game development is probably the same. Even if AI writes the code, I am the only one who can decide 'where to stop for today.' And it is the me of tomorrow who receives that continuation.

That is why, lately, I finish at 10:30 PM. There are days when I feel like I could do a little more. But I feel that rather than forcing myself to continue for that extra hour, it is better for the work in the long run if my tomorrow-self can resume in five minutes.

'That's it for today.'

Before, this was just a phrase to end my work. But now, I think of it as a phrase to pass the continuation on to my tomorrow-self.

A work of art is not something that suddenly comes into existence on the day it is completed. It takes shape little by little, as I stack up the days where I am able to receive the continuation from the me of yesterday.

Tomorrow, I will go to receive the continuation left behind by the me of yesterday. As long as I can receive that continuation, I believe the work has not yet stopped.

Recommended for you!