Modern benchmarks across the board really need to start severely penalizing disobedience and hallucination. A year ago models weren't really strong enough to justify this but they are now-- the frontier isn't in squeezing out the next bit of task completion, it's in making common cases not periodically be disastrously wrong.

A lot of the total cost of AI is fixing its "truth shaped errors", particularly in the presence of models that are very "gaslighty" when corrected.

GLM-5.2 is really the only model I've spent much time using that I didn't fatigue from being regularly lied to by the model, but that might be partially luck.