max @seoul.systems · 06/06/2026A recurring theme of lots of my thoughts around LLMs and their use in software engineering, is that their popularity points at least partially to the failure of our tools, our communities and lastly ourselves as engineers 120
max @seoul.systems · 06/06/2026Shell scripting might be the prime example here, with Bash being notoriously complicated and error prone. Another the insanely bad UX of most debuggers on Unix-like systems, and lastly the hostility of Stackoverflow (and derivatives) towards more novice members 030
