Re: Objection to the Doomsday Argument?

Nick Bostrom (
Mon, 17 Aug 1998 01:21:38 +0000

Doug Bailey wrote:

> Nick, I wanted to reiterate my request since I think
> it might have gotten lost in the thread. The problem I
> have with the DA is that it appears to me that any
> member of any species that applies the DA to its situation
> will always conclude that their species faces extinction
> in the near term.

You only get the near-term doom prediction if yo think population has been or will be increasing. With a constant population size, you should expect, other things equal, that the species will last about as long as it has already done.

>If Bayes' theorem is telling us only
> 99.999% of the species that analyze the DA should conclude
> imminent doom

If the DA is correct, everybody should apply it.

> but in practice 100% are then I begin to
> wonder if we are overlooking something, we're screwing
> something up in the application of Bayes' theorem, or
> drawing wrong conclusions.
> With that in mind, can you describe an individual that
> could apply to DA to its situation and not conclude
> doom was imminent?

For example, an individual who thought population were constant. Or an individual who had strong empirical grounds for not thinking doom imminent.

> What can we say about that individual?
> The items I would like to know are:
> What would such an individual know(i.e., about its species,
> its relative position, etc.)?

If she knew both her relative position and past and future population figures, then she could deduce exactly when doom would happen. So presumably she won't know her relative position.

> What is that individual's situation that would lead it to
> conclude doom was not imminent?

It has to do with that person's belifs about population figures and empirical risks.

> Could another individual sitting next to this individual
> reach a different conclusion?

Yes, if he had different information or different priors.

> Could another individual of the same species at an earlier
> or later point in time apply the DA and reach a different
> conclusion?


> (The answer would seem to be "yes" unless there
> was some reason to believe the species' existence was
> indefinite.) If the answer is "yes", what qualitative or
> quantitative differences would there be between the knowledge
> of the individual in question and the ancestor/descendant?

For example, a stoneage man might have thought there was some probability that the species might become extinct before 1000 AC. You are certain that that didn't happen.

Nick Bostrom
Department of Philosophy, Logic and Scientific Method London School of Economics