[extropy-chat] Role of MWI and Time Travel (was: Are ancestor simulations immoral?)

[Lee Corbin]

Samantha wrote

> On May 23, 2006, at 7:27 PM, Russell Wallace wrote:
> > On 5/24/06, Samantha Atkins <sjatkins at mac.com> wrote:
> > > I also suspect that one reason for creating a historical sim is to
> > > tweak the factors involved as minimally as possible to get a
> > > different and better outcome. This could be one way to learn more
> > > deeply from experience. 

That's an idea, yes, of fine SF pedigree, but not really compatible
with the Many Worlds Interpretation of Quantum Mechanics that you 
all go into below.

Russell:

> > There might be other reasons for doing it too.

> > Suppose you invented a time machine... What's the first thing
> > you do with it? along the way you stop off and kill Hitler.
> > If you subscribe to the many-worlds interpretation of quantum
> > mechanics (as I do)

Yes, me too.

> > then you have created an Everett branch in which Hitler died
> > early, therefore hopefully in which the Holocaust didn't occur.
> > This is good, is it not?

Samantha:

> As I understand it QM interpretations do not apply to macro
> level reality generally speaking. So I don't think MWI can be
> claimed to give you such a macro level branching. 

In the MWI, branching is occurring all the time, and everywhere.

I do think not that it is exactly correct to say that by making
a certain choice I am *creating* an Everett branch. It's much more
like I'm joining a certain pre-existing Everett branch. Remember
that QM under MWI is a completely deterministic theory, and that
"free will" is an awkward concept in deterministic systems.

So you don't prevent 60 million deaths in World War II by killing
the people who got the human race into it. That happens in a certain
definite fraction of the universes anyway.

Where free will is not entirely useless is when you decide to paint
your house green or brown. It feels like you get to choose, and 
indeed there are universes in which Samantha, Russell, or Lee
painted their house green, magenta, purple, yellow, and so on.
Some of these have very low measure, but their measure is vast
compared to extremely improbable universes (in which, for example,
a glass of water you were holding suddenly began to boil).

> > Now suppose you create a simulation of Earth ~1930 onward...
> > Given that by hypothesis a simulation is subjectively
> > indistinguishable from a "real" Everett branch, should this
> > not be considered good in exactly the same way?

Well, I don't know what distinguishes a "real" Everett branch 
from the one you are talking about. Time travel is theoretically
possible, and David Deutsch has a chapter on it in his great
"The Fabric of Reality", which is devoted to explaining MWI
and its philosophical underpinnings. 

> Hmm. I don't know. Do the beings within have the same chances
> for Singularity and transcendence? 

Unless there is a reason to suspect not---i.e., suspect a negative
correlation---then the beings everywhere have the same chances.
But it's true that younger people have better chances, and it's
true that other 1930+ universes have either a greater or less
chance of bestowing transcendence; for example, it may turn out
that World War II fostered technological gains that raise slightly
the measures of the worlds in which you and I gain transcendence.
But for all I know, those worlds that did not go through World
War II fare better technologically.

Lee