Granularity-Driven Development

　I’ve finished hacking all the equipment I have on hand. Since my funds are a bit tight, I’ve been working on ways to make better use of my existing peripherals for a while now. Once you clearly define what you want to do, all you have to do is give instructions to AI, and it’ll handle the rest. It all depends on how you give instructions to AI.

　The AI I’m currently using offers a free tier. Of course, if I had the money, I could pay to use it, but as I mentioned above, I don’t have the funds. So I’m sticking to the free tier—that’s right, I’m a free-to-play guy.”The fear of not paying is that you’ll hit the token limit before a task is complete, and it takes time for the tokens to recharge (on Claude code, which I use often, the recharge time is about 3 hours). On the other hand, if you break tasks down into small pieces, you end up in the absurd situation where you hit the limit even within those small tasks (which I consider “atomic”).

　Just like in any ordinary job, you’re expected to break things down to a logical stopping point. The key is determining the appropriate level of granularity for dividing tasks. Breaking tasks down along a timeline goes without saying. In addition, when tasks no longer interfere with one another, you divide them into parallel processes to reduce the workload of each task. This is granularity-driven development.

　There’s no point in regretting a lack of funds. There’s never enough money, storage, or time.