Ultimately I want my programming skills to produce value to others. I'll admit it's not something I cared about but the more I meditate, the more I want to be of service. It's also provided an extra gear in my drive to get better, understand as many aspects of programming as possible and in depth. At first I thought me pursuing my own interests are so vague and inconsequential that no one except me would find it useful. But being excited about what I am doing seem to be at the very least inspiring those around me and they in return inspire me with the things they have built, with tools and languages which are much different than my own. It's provided an answer to the question on some mornings when the alarm rings at 0400, "Why do this?." I'm glad I've stuck with it even when I did not have an answer equipped with only a desire to want to do it. A curiosity to see what would happen if I did this for a year. It's not a flex, not by the smallest means. I'm a very lazy person, I find it difficult to follow through with things except during the early hours of the day where I can effortlessly do things that I want. Even then I fall short some mornings. I picked 0400 because it seemed like a crazy idea to go to bed earlier than 2000. Now that I think of it, why is it so? Ah, I remember now, I don't plan on being single for long. Maybe I'll find someone who is an earlier bird than I am.
I'm quite lost most of the time. I don't know what I am doing or what I want to do next. The only thing I do know is that I want to program and do the best I can do. Even among this hype of AI I'm trying to force myself to curious about it. Not in the form of a prompt engineer, no thank you. I enjoy programming far too much for that. But rather using the building blocks of LLM's or Radial basis function to solve parts of problems which would have been impossible otherwise. I have a feeling that the current focus on productivity or replacing software developers (which may or may not happen, I have no fucking clue) is kinda thinking small. The example that comes to mind is AlphaFold, which found out all the ways in which protein structures could fold, or something like that. It's problem where the constraints are bounded but the combinations are high and tedious to find manually. I'm going to focus some of my efforts in solving or get better at solving such problems. This is quite exciting.
It's also why I think learning how to code is still valuable and I think it's only going to get more valuable. It's hard. Don't use AI as an excuse to proompt your way out of the difficulty. You are only digging your own grave. Proompt engineers will go out of business faster than anyone else. If there is an optimization that can be achieved with the input why won't Anthropic or OpenAI do it themselves. They want all engineers to be gone in about six months, do you think they'll keep the proompt engineer jobs around?