That's the analogy working as intended: the answer to "why do programmers still write memory-unsafe code" is the same shape as "why do microplastics researchers still wear gloves." The real answer is boring and full of tradeoffs. The HN thread version skips to indignation: "they never thought of contamination so ipso facto all the research is suspect"

(to go a bit further, in case it's confusing: both you and I agree on "why do people opt-in to memunsafe code in 2026? There’s no reason to" - yet, we also understand why Linux/Android/Windows/macOS/ffmpeg/ls aren't 100% $INSERT_MEM_SAFE_LANGUAGE yet, and in fact, most new written for them is memunsafe)