On Sat, Sep 15, 2007 at 01:25:04PM -0000, Rolf Nelson wrote:
> If I understand the Measure Problem correctly, we wonder why we find
> ourselves in a "Goldilocks Universe" of stars and galaxies rather than
> a simpler universe consisting solely of blackbody radiation, or a more
> complex, unpredictable Harry Potter universe.

I call this the Occam catastrophe in my book. The solution I give
there is a requirement that observers have to be embedded in the
universe they observe, ie are self-aware.

> 1. An attempt at the solution was that more complex universes are less
> probable; they are less likely to be produced by a random UTM. This
> explains why induction works, why we don't live in a Harry Potter
> universe. But this also means a simple blackbody radiation universe is
> more probable than a Goldilocks Universe.
> 2. So we say, "There are more observers in a Goldilocks Universe,
> where observers evolve through natural selection, than in a blackbody
> radiation universe, where observers can only occasionally emerge
> through extremely infrequent statistical anomalies." But if both the
> Goldilocks Universe and the blackbody radiation universe are infinite
> in size, then both have an infinite number of observers.

Unnormalisable measures are not an insurmountable problem. I give some
examples where this can be done in appendix C of my book. Of course
there are problems in the general case.


> Here is one possible solution: the UTM instead directly produces a
> qualia (or, if you prefer, substitute "observer moment" or whatever
> terminology you deem appropriate). We'll use a broad definition of
> "qualia" that can encompass complex observations like "Rolf sits at
> his keyboard, reflecting on past observations and wondering why he
> seems to live in a Goldilocks Universe", since that's exactly the type
> of observation that we're trying to explain when we ponder the Measure
> Problem.
> Each qualia, in the proposed model, is a long, finite-length string
> that is output by a UTM running every possible random program. (This
> is the same type of UTM that some of you have been proposing, but it
> outputs an attempt at a single qualia, rather than outputting an
> entire universe.) Very few strings are qualia; most UTM programs fail
> to produce qualia. The proposed model additionally postulates that
> many qualia are compressible in a certain interesting way, such that
> the World-Index-Compression Postulate (below) is true.
> World-Index-Compression Postulate: The most probable way for the
> output of a random UTM program to be a single qualia, is through
> having a part of the program calculate a Universe, U, that is similar
> to the universe we currently are observing; and then having another
> part of the program search through the universe and pick out a
> substring by using an search algorithm SA(U) that tries to find a
> random sentient being in U and emit his qualia as the final output.

This sounds kind of complex. Just how do you recognise sentience?

> As an example, take two qualia, that we will call Q(Goldilocks) and
> Q(Potter):
> Q(Goldilocks): "All my life I have read that all swans are white. And
> indeed, today I just saw a white swan."
> Q(Potter): "All my life I have read that all swans are white. But,
> today I just saw a black swan."

Funny you should say this - all my life I read that swans were white*,
but all the swans around here are actually black. It was only at the
age of 28 that I saw my first white swan - when living in Europe.

* in fairy stories of course - I knew full well that the first
  European exporers to our land were amazed at the black swans, and
  that they feature on the state flag where I grew up.


A/Prof Russell Standish                  Phone 0425 253119 (mobile)
UNSW SYDNEY 2052                         [EMAIL PROTECTED]
Australia                                http://www.hpcoders.com.au

You received this message because you are subscribed to the Google Groups 
"Everything List" group.
To post to this group, send email to [EMAIL PROTECTED]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at 

Reply via email to