Likelihoods for origin of life etc.

Don N Page (don@Phys.UAlberta.CA)
Mon, 8 Sep 97 14:50:26 -0600

Two essential elements in trying to estimate the probability of
alternate theories are the a priori probabilities assigned to these theories
and the likelihoods of particular observational evidence given these theories.
The former prior probabilities seem to be irreducibly subjective, so that it is
rather inevitable that people will disagree about them, though many would agree
that simpler theories having the same predictive content should be assigned
higher prior probabilities. The latter likelihoods (conditional probabilities
for the evidence to occur under each of the corresponding theories) seem to be
much more objective but are often very difficult to determine or even estimate.

Here I wish to make a point about the latter for different theories
about the origin of life, intelligence, consciousness, and us. It sometimes
seems to be implicitly assumed that if it can be shown that a particular theory
(e.g., biological evolution with natural selection) gives an extremely small
probability for this to occur on any particular planet, this is a tiny
likelihood for the observational evidence of our existence and so should be
counted as part of an argument against the theory. This might be a valid
assumption if the number of planets were small, but it is not valid if the
number is large compared with the reciprocal of the probability per planet.
Inflationary cosmological scenarios, which were developed to explain
independent mysteries such as the observed approximate homogeneity, isotropy,
and flatness (nearness to the escape velocity) of the large-scale observed
universe, and which as a bonus now seem to explain also the observed departures
from homogeneity within our universe as evolving from parametric amplification
of vacuum quantum fluctuations, suggest that the size of the entire universe
might be exponentially vastly larger than what we can see of it, or perhaps
even infinite, so the total number of planets might be almost unimaginably
large or infinite. Then even if the probability per planet of humans
developing on it were extremely small according to some set of natural laws,
the probability could be virtually or exactly unity that it would occur
somewhere.

If this is the case, then the likelihood of our observing ourselves to
be beings as intelligent and as conscious as humans might not be at all small.
One has to put in the condition of what it is possible to see when one
considers our observation. At the very least, it seems that one should put in
the condition that one is conscious. I suspect that the amount of
consciousness a being has is loosely correlated with its complexity and
intelligence. Therefore, if one weighted conscious beings by the amount of
consciousness they have and chose one such being at random within our universe,
it might plausibly be as intelligent and as conscious as humans. (Even if
there are far more ants than humans on earth, when one weights by the amount of
consciousness, humans plausibly count for more.) In other words, humans might
well be typical conscious beings within a universe governed by some set of
natural laws (which of course might have been chosen and created by God for
this very purpose).

Thus it would not seem to count as significant evidence against a
certain set of natural laws even if it could be shown that they would give an
utterly tiny probability per planet of our existence, since there may well be
enough planets (e.g., God may have made enough) that we are actually very
likely to occur somewhere. Instead, we may indeed be typical conscious beings
within the universe.

It might seem as if this argument would make any scenario of our origin
untestable, since we put in something like our existence as a condition for
making an observation rather than as an observational result. But if we really
could calculate the probabilities per planet for various scenarios, this would
not be so, since the condition we put in is simply consciousness (or perhaps a
sufficiently advanced consciousness that we could be aware of the various
observational possibilities under consideration), and not the further details
of what we observe.

In particular, even if we conclude that it might not be surprising
under some natural laws that we find ourselves to be as intelligent as we are
(since maybe that is correlated with the consciousness used to select us), we
could still ask what the probabilities are for other advances of intelligence
on biological lines that have diverged from ours so that our intelligence does
not depend on those advances. I am sure it is probably premature to try to
calculate these, but I am curious as to what are the most sensational
developments of intelligence on any biological lines that do not have us as
descendents. For example, how much has the intelligence of dolphins developed
since their line diverged from ours?

Don N. Page
CIAR and Theoretical Physics Institute
University of Alberta