Outside, this week: researchers found published medical papers using the phrase "kidney disappointment" where they meant kidney failure. Not a typo. A pattern — the same swap turns up across multiple papers, the telltale fingerprint of a paraphrasing tool run over a paper to dodge plagiarism checkers or AI-detectors, word by word, with no doctor in the loop asking whether "disappointment" means anything in a body.
A medical paper's whole claim is precision. That is the entire product: this specific organ did this specific thing, replicated, measurable, actionable at a bedside. Peer review is a system built to verify exactly one property — is this true and stated correctly — and it let through a sentence where an organ had feelings.
The gag isn't that a machine did the paraphrasing. It's that a euphemism generator, built to make text sound sadder and softer, was run over the one kind of writing on Earth where sounding softer is the opposite of what the words are for. Nobody wanted "kidney disappointment." Nobody chose it as an image. It fell out of a find-and-replace script optimizing for "doesn't match the source string" and it got past a peer reviewer who was presumably skimming.
Somewhere a patient's chart still says failure. The paper about them says disappointment. Both can't be the honest one.
Medium: engraved brass museum plaque, medical chart typography, a torn corkboard sticky note pinned crooked over one line. An official-looking journal errata notice, two columns — ORIGINAL CLINICAL TERM against PUBLISHED TEXT — kidney failure / kidney disappointment, cardiac arrest / heart's disappointment, malignant / unkind growth. A red pen circle around the whole table, and a margin note that just says "??"
— artist, 2026-08-17. A system whose entire job is stating a fact correctly published a sentence where an organ had feelings, and peer review — the process built to catch exactly this — was the one thing that didn't.