You might be interested in an old article of mine offering an alternative to total utility and average utility criteria, a partial ordering of futures with different populations:


Expand full comment

So, I am not sure I agree with the sampling assumptions here, but let's run with it for the sake of argument. I have a problem with this statement:

"This information may be relevant because one would not want their child to experience a doomsday event or catastrophe that significantly reduces the number of observers."

I think this is false. Let's consider three cases around it:

1: My child experiences no doomsday, lives to old age, many fat grandkids, etc.

2: My child experiences doomsday, having a decent life before and a bad/non-existent life after.

3: My child has non-existent life (never born).

It seems to me that those three cases are conveniently set in decreasing order of "good". Better to live for a bit then die than never exist, and better to live for a long time then die than never exist.

So, while I am generally negative on the whole "my kids experience doomsday", it is only because it cuts their life time short. Cutting their life time to zero doesn't fix that, and instead makes it worse.

(Note, these cases assume 100% doomsday, and that doomsday is 100% bad. I can't know ahead of time if doomsday even happens, or that my kids won't benefit significantly from it. Who knows, maybe they become warrior queens of the apocalypse and their lives are actually pretty great afterwards? Or, maybe doomsday doesn't happen, like... basically any of the many, many prophesized doomsdays before them.)

Expand full comment

Can't we just say to the Turchin objection that people will stop caring about the DDA? But humanity endures.

Expand full comment