Could you just have superhuman Go AI just how good humans are somewhat more objectively?

Not without flaws of course, but probably interesting