Re: Fermi Paradox in the news

From: Nick Bostrom (
Date: Thu Oct 26 2000 - 23:45:18 MDT

Jason wrote:

>At 03:58 PM 10/26/2000, you wrote:
>From: "Nick Bostrom" <>
> > Yes, but it would not be necessary that we would have found ourselves
> > originating from a solar system where two civilizations had evolved
> > independently. That is the extra finding that would make all the
> > difference. If the evolution of intelligent life is very improbable then
> > would have been extremely unlikely that it would have happened on another
> > planet in our own solar system.
>No, you see, you're stuck measuring things on a plantary scale. Couldn't we
>instead surmise that our -solar system- was particularly good at creating
>life-bearing conditions?

Yes, if you thought that the improbability lay in the task of getting a
suitable solar system together and that once you had the right
constellation of planets etc then it would be easy for organic matter to
emerge and evolve into intelligent creatures. However, empirically that
looks highly implausible. We have already discovered over fifty extra-solar
planets and have every reason to believe that the universe abounds with
them. Stars that are similar to our sun are also very common. By contrast,
there are many excellent candidates for the great filter between "there
being a suitable solar system" and "intelligent life has evolved".

The one possible qualification to this is if you think life might have
jumped from one planet to another as a result of a meteor or asteroid
blast. But if we're talking about discovering life that's evolved
independently on another planet in our vicinity, then that's very strong
evidence that life is quite common in the universe.

>Let us take a step back. Let's pretend that planets don't develop life:
>realities develop life. (Or rather, realities develop universes, universes
>develop planets and planets develop life.) Let's further pretend that "we"
>doesn't equal humanity, but rather "we" equals "all observers." This is
>what the AP really says after all, doesn't it: i.e.: Observers should not
>be surprised to be observing a reality that begets observers. Right?

I'm not keen on entering a debate about what "the anthropic principle"
says. I've counted over 20 different and incompatible definitions in the
literature - it's simply a big mess! But the interesting idea, the one that
motivated Brandon Carter who invented the term "anthropic principle", is
that of observational selection effects. A first approximation of this is
given by the Self-Sampling Assumption, which says that: "Each observer
should reason as if they were a random sample from the class of all
observers in their reference class." (Please see my writings for how this
is to be interpreted.) I see this an a deep (and non-trivial)
methodological principle, for which compelling arguments can be given. It
has lots of interesting consequences. This is an exciting new area, where
philosophers and physicists are cooperating in fruitful ways. When thinking
about the large-scale structure of the world and the place of observers
within it, observational selection effects must be taken into account. It
also has applications in down-to-earth fields, like traffic analysis.

That reality begets observers might well be surprising. It depends on what
reality is like. If things were set up in such a way that it would have
been extremely improbable that any observers should have come to exist
(imagine e.g. some cosmic fortune wheel that had to stop in a very unlikely
spot for life to emerge) then our existence would indeed be surprising. It
is partly in order to avoid stipulating such a surprising contingency that
people are leaning towards theories (such as multiverse theories) according
to which it would be much more probable that observers should exist.

Nick Bostrom
Department of Philosophy
Yale University

This archive was generated by hypermail 2b30 : Mon May 28 2001 - 09:50:18 MDT