Could you or someone explain what exactly this means? My interpretation is that he's estimating based on the number of ways you can rearrange existing genes floating around in the aggregate human genome that it is possible for humans to exist who are 30 standard deviations smarter than average. This violates my heuristic that one should avoid extrapolating far outside the domain of actual data.
This is very interesting, please tell me if I'm misunderstanding something.
That was my understanding as well. As discussed in the comments, SDs/IQs are just percentiles - ordinal, not cardinal -, so it's not obvious what 30 SDs would translate to. If Einstein is 15 SDS out, maybe 30 SDs just means you are a physics genius like him but also a little bit wittier than he was.
If it's worth saying, but not worth its own post, even in Discussion, it goes here.