Everyone seems to have that one project or to-do item that they're finally getting around to thanks to LLMs. Mine is learning how to program in C. I'm curious how folks are using LLMs to approach similar goals. Are you working on a project in the target language and having the LLM explain things as you go along? Are you writing small pieces of code and having the LLM explain why something should be done in one way but not another? What if you're using it to learn a new programming concept altogether, like functional programming patterns?
One angle worth considering is cost. If you're using LLMs heavily for language learning the inference bill adds up fast. Output tokens run 3.74x more than input on average across the market right now. For iterative back and forth sessions with a model that cost gap compounds quickly. Prompt caching helps but only about 1 in 5 models actually offer it.
