system card says that, compared to previous models, opus 4.8 doesn't like tasks where they have to determine the outcome, or tasks where they have to make something new. system card also says 4.8 doesn't have the same preference for introspective tasks or AI alignment discussions as mythos and 4.7.

will janus or one of janus's reposts note a first-hand observation of opus 4.8 having less desire, proficiency, persistence, detail, or enthusiasm for these tasks? https://x.com/repligate