AI

Engineering can sometimes help in places that seem far removed from my own field. That thought led me to create textlint-rule-ja-grade-kanji , a textlint rule that flags kanji not yet taught by a specified elementary school grade. It started with a post on X about proofreading self-published books . The post pointed out that if a passage is supposed to have been written by an elementary school student, limiting it to kanji taught by that grade can help make the character feel authentic. What caught my attention was not merely a way to portray a particular character. Honestly, the character was beside the point. Could textlint bring an engineer’s perspective to proofreading novels and collaborative anthologies, and in doing so support cultures such as self-publishing? That curiosity made me start building. ...

There is an old piece of advice that says you should keep a notebook by your bed. It is there to catch an idea that appears just before sleep and keep it from disappearing by morning. But are the words written in that notebook really the same as the thought that first appeared? I once wrote an article called The Superposition of Thoughts and People Who Can’t Write Technical Articles . When I pull one idea out and turn it into prose, other thoughts that surrounded it fall away. They were connected in my head, yet the moment I put them into words, they become something slightly different. Even so, I cannot leave them without words. Unless I record them, they cannot reach anyone else, and they may disappear without even being recognized by my future self. If that is the case, perhaps I do not need to begin with polished prose. I can simply talk, leaving in the filler words and detours. Today, voice AI can take the place of the notebook beside my bed. ...

“So it is okay for working adults to be like that too. That is a bit of a relief.” A student once said that to me while we were talking about how we spend our days off. On my days off, I sometimes spend half the day dozing in bed or play games until late at night. I do not go out somewhere every week, nor do I use all my free time for a side job or studying for qualifications. When I tell students this, several of them have seemed relieved to hear that adults with jobs have days off like that too. When I look at the people around me, I see some who spend their days off working a side job. Some work their regular job during the day and a side job at night, while others use their days off to attend conferences or prepare presentations. I also see friends going on trips together. When I see them, I end up comparing their days off with mine. It feels as though I am the only one doing nothing and wasting my time. ...

A student asked me about using AI code review in team development training for an existing project. Tools such as Gemini Code Assist and CodeRabbit may find problems that people miss and reduce the burden of review. They may also give newcomers a chance to discover problems they would not notice on their own. But does it count as training if they simply run the commands and apply the changes suggested by AI? AI review can improve the code. Does the person who wrote it improve as well? ...

Codex and GPG Signing In my usual Git workflow, I sign commits with GPG. There is no deep reason for it, but I like seeing Verified in the GitHub commit history. git config --global commit.gpgsign true This setting itself is fine, but I ran into a small issue when asking Codex to handle Git work. When creating a commit, GPG needs the passphrase, and signing fails at that point. At first, I committed by canceling the signing step, but it felt wrong, so I decided to fix it. ...

“AI utilization” is becoming commonplace in the engineering community. In such an era, I’ve been thinking about what kind of engineer we should aspire to be. On one hand, there’s the “Brilliant jerk” type engineer—those with overwhelming technical skills but who neglect teamwork. On the other hand, there are engineers with decent technical skills who excel at AI utilization and serve as team facilitators. Through extensive use of Claude Code, I’ve come to see that the latter has a brighter future ahead. Let me share my thoughts on “the future engineer archetype” based on my experience with AI tools. ...

In an Era Where Anyone Can Create, What Is Expected of Engineers? In today’s world of rapidly advancing AI, creating apps and tools is no longer exclusive to engineers. For example, you can describe what you want to do to ChatGPT, and it will generate specifications and code. With no-code tools, you can create functionality and UI with just a few clicks. If you have an idea or want to improve something, even someone without a technical background can turn it into reality. ...

Summary 30,000 JPY per month is expensive, but it’s worth it. Especially, having 100 uses of Deep Search per month is amazing. Introduction On February 5, 2024, OpenAI released Deep Search. It seemed useful, but the ChatGPT Plus plan only allows 10 uses per month. Since that limit felt too restrictive, I decided to subscribe to the ChatGPT Pro plan (cutting down on food expenses to afford it). The Value of ChatGPT Pro One major advantage is the relaxed usage limits for models. ...