Comment on Top AI models fail at >96% of tasksparentComments−rsynnott7moI mean performance is so bad across the board that this is likely essentially random. Monkeys accidentally doing a bit of Shakespeare.−ben_w6moThat's wildly overestimating what monkeys can do on a typewriter.It takes a lot to just be mediocre. Which, don't get me wrong, I'll agree current ML is, it's just that "mediocre" is an incomprehensibly huge step up from "random".
Comments
I mean performance is so bad across the board that this is likely essentially random. Monkeys accidentally doing a bit of Shakespeare.
That's wildly overestimating what monkeys can do on a typewriter.
It takes a lot to just be mediocre. Which, don't get me wrong, I'll agree current ML is, it's just that "mediocre" is an incomprehensibly huge step up from "random".