- Contemporary messages sorted: [ by date ] [ by thread ] [ by subject ] [ by author ] [ by messages with attachments ]

From: Jacques Mallah <jackmallah.domain.name.hidden>

Date: Tue, 16 May 2000 18:53:08 -0700 (PDT)

--- Jacques Mallah <jackmallah.domain.name.hidden> wrote:

*> Measure, basically, is the number of observers
*

within

*> a set. Thus if almost all observers were supermen
*

*> within the plenitude, us being regular Joes would
*

*> falsify the TOE.
*

On second thought, it wouldn't. The problem (if

such it is) is that Occam's razor gives a huge initial

favoritism toward the AUH. Any observation is

compatible with both the AUH, and some more

complicated hypothesis B. According to Occam's razor,

the initial Bayesian prior to take for B is just equal

to the *same* as the Baysian probability that we would

see the things predicted by B even if the AUH is true.

Thus we would always conclude that either B is true,

or the AUH is true, with about equal Bayesian

probability.

The only exception is that anthropic

considerations tend to favor the AUH.

In the above example, suppose (hypothetically!)

the AUH predicted that only 1 observer in 10^100 is

_not_ a superman. The AUH thus very strongly

predicts, we should be supermen.

But suppose the Kolmogorov complexity of B (the

hypothesis that most observers are NOT supermen) gives

it an Occam weight of 10^-110. Then we would *still*

conclude that there is only a 1 in 10^10 chance that

the AUH is false.

Only if the fraction of observers, in the AUH,

with some characteristic is smaller than the Occam

weight of an alternative theory that only allows such

observers would we conclude that the AUH had been

falsified. That is practically impossible.

One way around the problem is to be prejudiced

against the AUH. Then we could argue that for certain

observed characteristics, we will only believe the AUH

if the AUH predicts a large enough effective

probability to have that characteristic. I think most

people would agree with that approach. I would like

to make this more precise though.

On the other hand I don't think falsifiability is

as important as it's cracked up to be.

=====

- - - - - - -

Jacques Mallah (jackmallah.domain.name.hidden)

Physicist / Many Worlder / Devil's Advocate

"I know what no one else knows" - 'Runaway Train', Soul Asylum

My URL: http://hammer.prohosting.com/~mathmind/

__________________________________________________

Do You Yahoo!?

Send instant messages & get email alerts with Yahoo! Messenger.

http://im.yahoo.com/

Received on Tue May 16 2000 - 18:55:42 PDT

Date: Tue, 16 May 2000 18:53:08 -0700 (PDT)

--- Jacques Mallah <jackmallah.domain.name.hidden> wrote:

within

On second thought, it wouldn't. The problem (if

such it is) is that Occam's razor gives a huge initial

favoritism toward the AUH. Any observation is

compatible with both the AUH, and some more

complicated hypothesis B. According to Occam's razor,

the initial Bayesian prior to take for B is just equal

to the *same* as the Baysian probability that we would

see the things predicted by B even if the AUH is true.

Thus we would always conclude that either B is true,

or the AUH is true, with about equal Bayesian

probability.

The only exception is that anthropic

considerations tend to favor the AUH.

In the above example, suppose (hypothetically!)

the AUH predicted that only 1 observer in 10^100 is

_not_ a superman. The AUH thus very strongly

predicts, we should be supermen.

But suppose the Kolmogorov complexity of B (the

hypothesis that most observers are NOT supermen) gives

it an Occam weight of 10^-110. Then we would *still*

conclude that there is only a 1 in 10^10 chance that

the AUH is false.

Only if the fraction of observers, in the AUH,

with some characteristic is smaller than the Occam

weight of an alternative theory that only allows such

observers would we conclude that the AUH had been

falsified. That is practically impossible.

One way around the problem is to be prejudiced

against the AUH. Then we could argue that for certain

observed characteristics, we will only believe the AUH

if the AUH predicts a large enough effective

probability to have that characteristic. I think most

people would agree with that approach. I would like

to make this more precise though.

On the other hand I don't think falsifiability is

as important as it's cracked up to be.

=====

- - - - - - -

Jacques Mallah (jackmallah.domain.name.hidden)

Physicist / Many Worlder / Devil's Advocate

"I know what no one else knows" - 'Runaway Train', Soul Asylum

My URL: http://hammer.prohosting.com/~mathmind/

__________________________________________________

Do You Yahoo!?

Send instant messages & get email alerts with Yahoo! Messenger.

http://im.yahoo.com/

Received on Tue May 16 2000 - 18:55:42 PDT

*
This archive was generated by hypermail 2.3.0
: Fri Feb 16 2018 - 13:20:07 PST
*