A sepsis model scoring 0.63 AUROC caught one case in three at the threshold hospitals used
The area under the ROC curve scores how often a model ranks a sick patient above a well one, a measure that stays silent on the threshold a hospital sets and on how rare the disease is.
Hidden reasoning went from a built demo to a training byproduct
A hidden channel had to be hand-built on a rigged task, but reasoning that turns illegible on its own is the harder problem for anyone who oversees AI by reading what it writes.
AlphaZero beat Stockfish twice, under different rules each time
Everyone quotes the one-minute-per-move match from the 2017 preprint, not the peer-reviewed 1000-game rematch that gave Stockfish tournament time and still ended 155 to 6.
Image generators put a prompt's colors on the wrong objects
On the standard benchmark, Stable Diffusion binds a color to the right object at 0.38 out of 1, because the prompt is encoded with no record of which adjective belongs to which noun.
Two Waymo cars hit the same towed truck minutes apart
A pickup towed backwards and angled across a turn lane was too rare a shape for the step that predicts where other vehicles will go next.