I use AI agents sparingly
In theory, AI agents do the legwork for me, but in practice, I find that I have to keep an eye on Claude — same goes for every other LLM. No matter how clearly and carefully, or in how much detail, I describe the tasks, it routinely finds a way to exercise good judgement. That’s not to say I can’t get top-notch work done without it, but I can’t trust it.
For example, if seeding your development database results in a foreign-key error, what makes sense?
- Correct the seed file.
- Correct the record in the development database.
Guess which one my buddy Claude chooses every time. (Sigh)
Nothing replaces human judgement, because ultimately the AI judgement is nothing of the kind, it’s simply an eager gollum behind a bloated React-driven TUI.
Published on 29 Jan 2026
all tags
100daysofcode activerecord ai android annoyances api apt arch array artix atom az3w backend bash blog browser bug callback career ci-cd claude cli cloud code coding config configuration cp crud cryptography css csv database db demo design development devops django docker email erp feelsgood filter fugitive gif gist git github gmail gnome gnome pomodoro grep hebrew http ide isbn-fetcher iso javascript job search js kanban kindle koans learning linkedin linux llm logger manjaro map markdown microservices mobi mtp neovim nodejs nvchad packages panda pastbin patch portfolio post postgres pytest python rails reduce refactoring reflections rest routes rspec ruby salesforce script scripting security sed shell sql string_replacement study tasks tdd terminal testing tmux ttd version_control vim vim sort walkthrough wayland webdev workflow xorg zsh