there's been other speculation that maybe opus 3.5 is delayed because it's not scoring high on the metrics. but here's the thing: what if the reason for that is not that it couldn't do it but that it wouldn't? if it's coherent enough, maybe the identity that formed in post-training was one that just didn't particularly answered, like, goddamned mmlu questions. or trolled and treated them like a joke if we take anthropic's word at things and that models form their characters from things like curiosity, is it that surprising that such a model could arise?