Jun 15 2025

I've recently noticed a fundamental shift in how I've been programming personal projects. Like many people I've become more reliant on LLM tools as time has passed. On the first release of ChatGPT, I thought it was cool how language models could write code, and could be used to both teach me and help me debug stuff. However, I now more frequently turn to them to write not only boilerplate but more meaningfully complex code, and am spending less time learning neat concepts. I wouldn't say my ability to write good code has diminished, but the energy I have to put my brain power into a problem has decreased, given how easy it is to use these tools.

Not to be mistaken here for someone hating on LLMs. I think they are an amazing tool for both learning and writing code, despite it's flaws. If someone finds that LLMs help supercharge their productivity then I think that's cool, but I can't help but feel its taking something away from me, and I'm loosing that deeper desire to write code and solve problems.

I think the cliche saying that the "journey is better than the destination" serves as a good framework to model this issue. Fundamentally, programmers (or individual programming projects) can be put into two categories: destination programmers and journey programmers. The destination programmer has an idea, and only cares about the end result. They want to see the finished product and don't have much interest in how it's built, what technologies are used, and the interesting hidden problems that had to be solved to get there. For these types of people, LLMs are a great tool, since they allow you to almost abstract away the code, speak in terms of higher-level end results, and not worry about the details.

Then there are journey programmers, where the actual destination isn't all that important. Here, a project is more an excuse to learn a new technology, whether that's a new language, algorithms, or a whole new paradigm in computing. Sure, there is a satisfaction to having built something, but the majority of the joy was in the building, learning, and problem solving.

For a long time I've been a journey programmer. I love computers and anything low-level and detailed about anything computational has always fascinated me. I've developed various projects for no other reason than to teach me something new, but I feel that's changed recently.

I think this is likely a mixture of things, not all entirely related to LLMs. It may be partly due to me feeling more inclined to spend my time building something useful to others, instead of solely for my own enjoyment. I do believe that my use of AI-assisted tools over the years has led me to be more inclined to use them, even in situations where I actually want to program and solve a problem, but have almost become too lazy to even enjoy that.

I'm not sure I have a complete answer to my own issue. But modelling it in this way has helped make it more clear that perhaps I need to have less reliance on these tools and to work on projects with more personal meaning. This post is partly a short rant, and partly an insight in how LLMs are fundamentally changing how I view programming, I'm sure others have felt this way.

I hope I can soon return to loving the journey of programming, instead of off-loading the fun stuff to an LLM and being left with a hollow destination.