I seem to be having more success (and fun) using agentic AI (locally with small models) than I have had with trying to use larger models for editing code. I think there is a major reason why: I tend to enjoy coding (at least at times), and I tend to dislike managing websites.
It makes sense that using agentic AI—running llama-cpp
with a small local model like Gemma 4 E4B and an MCP server—is more enjoyable than doing an unpleasant task. So it seems like the real benefit here might be minimizing the amount of time I do unpleasant things.
Some people might dislike coding (and I don't blame them) but I have always found it to be interesting—although exhausting mentally. However running command lines to update files and such on a website has always been boring and rote for me.
It is probably worth putting effort into trying to use agentic AI (like building an MCP server for local use on llama-cpp
) with tasks that are unenjoyable. For things that one finds enjoyable, agentic AI may have less utility.