Re: consciousness based on information or computation?

From: <>
Date: Wed, 27 Jan 1999 17:29:42 -0800

I wrote:
> I find it hard to accept the premise that "I" am a random selection from
> among all possible conscious/intelligent entities.

Wei replied:
> Why do you find it hard to accept?

It is difficult for me to understand what this premise would mean.
I know that I am a human being in a certain place and time. How can
I say that I might be, or might have been, a poor Chinese peasant?
An inhabitant of ancient Rome? Or a bug eyed alien?

The one way I can make sense of this is a rather weak reading: if all
conscious entities try to draw a conclusion based on the assumption
that they are randomly selected from among all conscious entities,
then most of them would be right. Suppose it turns out, for example,
that most conscious entities exist in groups near others, rather than
in isolation. Then for most conscious entities, the conclusion that
"I am near others" would be true.

This is basically tautological. It is just a matter of the definition
of what "most" means. Things that are known to be true for most observers
are true for most observers. But I have the impression that we want to
make a stronger statement here.

This leads to the problem of what is the proper reference class. If all
I know is that I am a conscious observer, then I can draw these kinds
of conclusions. But I know more. I know that I am human. If I know
that humans have hemoglobin, then I can conclude that I have hemoglobin.

I know more than this. I know that I am a human being living in the
year 1999, a resident of the U.S. From this I can conclude more about
my likely circumstances, for example, if I have a heart attack I can
probably get medical treatment in about 15-30 minutes.

But it seems to me that I know even more than this. I know that I am a
specific individual, with specific memories and personality. This is the
maximal information I have available to me. The only actual uncertainty
I have about my position in the universe must be consistent with this

There may be other instances of consciousnesses which are
indistinguishable from mine somewhere in the universe. I don't know
which of those instances I am. It seems to me that this is the proper
reference class from which I should consider myself a random selection.
It is the only one which fully takes into consideration the information
I have available.

> Of course you can choose not to accept
> it, but then you lose a very important advantage of the "everything"
> theories, which is an explaination of why our experiences are so simple
> and regular, as opposed to random and arbitrary. How would you explain
> this fact without the above premise?

If we could show that among instances of me, most were in lawful
universes, that would be a good first step. It would not show why I
existed in the first place, but it would give me reason to conclude that
the apparent lawfulness of the universe was not an illusion.

I earlier wrote:
> In this way I do not try to draw conclusions about the likelihood or
> longevity or measure of other consciousnesses than the one I am now
> experiencing. I don't conclude that the human race is doomed, or that
> I won't live forever, or that few aliens exist. These conclusions are
> very questionable, in my opinion.

And Wei replied:
> Me too, but I think the problem is in the arguments, not in the premise.
> If you care to exhibit such an argument, I'll try to point out the error.

The doomsday argument has a large amount of literature on it. Nick
Bostrom has a page at
with several links to it. These people have spent a lot more time
thinking about it than I have, and if there is an error, no one seems
to agree on it. Nick and Robin Hanson had a long debate about the
subject on the extropians list a few months ago, but were not able to
come to agreement. (They seemed to bog down on the issue of the proper
reference class, whether it made sense to say that I might have been
a rock. That was what I got out of the debate, anyway.)

Received on Wed Jan 27 1999 - 17:39:53 PST

This archive was generated by hypermail 2.3.0 : Fri Feb 16 2018 - 13:20:06 PST