For ten years I've been disciplining myself to avoid "snacks", small, easy, fun tasks that people use as excuses to avoid solving difficult problems. Suddenly, snacks are free because of LLMs. But the difficult problems remain.

I'm cautiously optimistic that this is the tool that will allow me to start coding with AI agents for real. I have a very one-track mind, which makes running multiple branches of a project really hard for me. Having a kanban-style interface keeps me from feeling completely overwhelmed and confused. Plus, I can ignore it and code by hand while I wait for it to finish. View quoted note →