That's a good point - the distribution over time should affect the answer. I've heard this rule of thumb applied before to the same problem in a different way

That's a good point - the distribution over time should affect the answer. I've heard this rule of thumb applied before to the same problem in a different way - rather than assuming you're somewhere in the middle of the time period humans exist for, assume you're somewhere in the middle of the list of all humans in order of birth. (This version probably makes a much more pessimistic prediction than the article does?) Edit: tgb got there first :)