on llms

a short note on how i use llms when building and writing.

everything on this site is mine. the ideas, the projects, the opinions, and the things i get wrong. all of it starts with me sitting down, trying to understand how a system works, and writing code or notes until it makes sense.

i use language models regularly. to me, an llm is just another tool in the workflow, like a compiler, a debugger, or an editor. specifically, i might use one to:

• fix typos, awkward phrasing, and punctuation i stopped noticing after drafting
• help format files, site templates, and css markup
• bounce technical ideas around or catch edge cases i might have overlooked
• tighten up a paragraph when it gets too wordy

what an llm does not do is decide what i think or what i build. i don't generate empty filler articles, and nothing goes on this site without me reading, understanding, and standing behind every single line.

using good tools doesn't need an apology, but being upfront about your stack always matters.

inspired by nisarga.