On Fri, Jan 30, 1998 at 10:12:22AM -0800, Hal Finney wrote:
> So, the idea is that there is some probability that your brain could
> form spontaneously in intergalactic space, and that probability per
> brain-volume would be approximately 2^-L?
No, that wouldn't work because the density of interstellar gas is too low.
The idea is that if you look at an appropriately larger volume of space,
there is some probability that the gas particles in it are lined up so
that they have the same basic structure as WDB (i.e. the organic brain),
even if they are much farther apart and are made of different atoms (these
differences do not matter since they can be removed by simple
transformations). This probability is approximately 2^-L. If you're
worried about the encoding scheme L is defined in terms of, let's instead
define L in terms of this probability. It seems any reasonable encoding
scheme should encode WDB in not much more than L bits.
I want to make two points:
1. The probability of finding WDB in a random brain-volume is much more
than 2^-L. We should be able to derive this from quantum mechanics, but my
argument is that otherwise I would have to believe I'm likely to be
instantiated as interstellar gas.
2. The number of branches in the universal wavefunction is enormous (see
Q11 of the many-worlds FAQ). It's much bigger than 2^L. The conclusion
from this is that I and everyone else must be living in multiple branches
simultaneously, and many of these branches do not share any recent
history (i.e. they became decoherent early in the history of the
This implies that when someone observes a new piece of information,
sometimes that causes a split, but other times it merely causes a
differentiation. So many-worlds and many-minds interpretations are each
correct in different situations.