Day 14 - Luckily, thinking does not have to be done in front of a computer

I “finished” my spec early morning. My intention was to get a basic knowledge graph working so that I could get an idea for how to build out a recommender system.

I had used neo4j before and fed if a bunch of flash cards. I visually saw the relationships and how easily they formed with new data. I knew there was something “there” without knowing what the end goal looked like.

And so I was interacting with both Claude and Gemini about the topic and they were both like “why the hell would you use a knowledge graph without a recommender system defined from the start”. I couldn’t get it through “their heads” that it was a discovery activity for me.

Nevertheless, throughout multiple conversations, I thought of an algorithm to start.

But it took me the entire day to think about it. Thinking is difficult work especially when you have no idea of the problem you have. The obvious answer is to just come at it from a different standpoint, but this is mentally taxing. To work on something and then reach a point and say “idk actually, I guess I need to start over a bit”. The good thing is that you don’t have to be sitting in front of a computer to “think”. You can go out and enjoy your day and take breaks to “think” - the ideas end up appearing for me. I like to think that my subconscious is making a soup - I don’t know when it will be ready, but I’ll be able to smell when it’s done.

So that’s what I did today. Got stuck, went to the gym, had some food, watched a judo class, and then made my way back home. And in-between these activities, I read through my spec, had different conversations with Claude and Gemini, and got closer to a v1 recommender system.

P.S. pic is from the judo training I watched - it’s open to the public and required before being able to enroll.