Satan Oscillate My Metallic Sonatas

tronvillain, Malaclypse:
How is this a problem?

I hear you and Malaclypse SAY it is a problem ...but you don't SHOW that this is a problem.

Does even really matter WHEN I know something completely independent of me? Does this change the probability?

I am certainly willing to listen to your thoughts on the matter.


Note: It seems to me that to show this is a problem one MUST deduce that the probability of God creating the universe AFTER He created the universe is different then BEFORE He created the universe. However, this is completely absurd.


Thoughts and comments welcomed,

Satan Oscillate My Metallic Sonatas

Malaclypse the Younger

Tercel

Nothing wrong with that. And I have no problem with simply stipulating that P(E|D) above is on the close order of unity. It is possible to quibble over this number, but it's not where the meat of the argument lies.

Generally speaking when there's 10 orders of magnitude difference in the range of one of the assumptions (especially one of the assumptions that's peripheral to the argument), this is a giant clue that numbers are being pulled out of orifices better left unmentioned.

The point is that we have no rational basis for believing anything about what a being capable of creating universes would or would not be likely to do.

As I said, this point is a quibble. Let's stipulate it's on the close order of 1 (which 0.5 certainly is), and move on to the meat of the argument.

Without making some assumptions, it's impossible to assign it any rational value whatsoever.

And now we come to the meat of the argument.

What does or does not impress you is irrelevant to a rational discusion. And quit with the "million million million", it's needlessly rhetorical. Stick with exponential notation: 10^-18.

It is stipulated that the a priori probability of any particular universe arising by chance is 10^-50.

If you arbitrarily assign a value for the a priori existence of a designer higher than 10^-50 (+/- the couple of orders of magnitude you throw in to the peripheral assumptions), you will infer design. If you arbitrarily assign a value lower than 10^-50, you will infer chance. If you assign a number equal to 10^50, you will find yourself uncertain.

To assign a value to P(D) greater than 10^-50, you are assuming that it is a priori more plausible to believe in the existence of a designer than to believe that the universe arose by chance. You can sanitize that assumption with multiplication all you want, but it is unsurprising that if you assume that a designer is more probable, you will conclude that a designer is more probable. This is the sine qua non of circular reasoning.

Yes, you get out the 50 (actually 20) orders of magnitude you put in with your assumptions. Quelle suprise. Hiding your circular reasoning with excess verbiage and several different methods of representation doesn't really help much.

[ February 25, 2002: Message edited by: Malaclypse the Younger ]</p>

tronvillain

SOMMS:

Quote:

As such, a more accurate analogy is the situation in which you have one 'fair' die and one '1093-biased' die. You choose a die (you don't know which) and roll a '1093'.

With the dice, you claim to already know that of them is biased towards 1063. If you roll one of them and it comes up 1063, it is extremely good evidence that it is the biased die.

However, what if you only know that one of them them is biased towards some number? If you roll one of them and it comes up 1063, it is not evidence that it is the biased die at all.

I'll show you:
If h is "This die is biased" then P(h)=0.5, P(~h)=0.5, P(e/h)=1/10^12, and P(e/~h)=1/10^12.

P(h/e)=P(h)P(e/h)/[P(h)P(e/h)+P(~h)P(e/~h)
=(0.5)(1/10^12)/[(0.5)(1/10^12)+(0.5)(1/10^12)]
=0.5

[ February 25, 2002: Message edited by: tronvillain ]</p>

tronvillain

Now, let's look at the hypothesis "This die is biased towards 1063." For this h, P(h)=(0.5)(1/1^12), P(~h1)=(0.5), P(~h2)=(0.5)(1-1/1^12), P(e/h)=1, P(e/~h1)=1/10^12, P(e/~h2)=0.

So P(h/e)=(0.5)(1/1^12)(1)/[(0.5)(1/1^12)(1)+(0.5)(1/1^12)+(0.5)(1-1/1^12)(0). Now, if you'll do the math, this is equal to 0.5! Of course, P(h/e) is much larger than P(h), but it is exactly equal to the probability of the die being biased! You have simply narrowed the possible states of the world - either this die is not biased, or it is biased towards 1063. If you roll the other die, you will be able to say essentially the same thing about it.

Satan Oscillate My Metallic Sonatas

tronvillain,

Thanks. I must sleep now, but I look forward to thinking about this tommorow.


Will most likely have some more questions.



Satan Oscillate My Metallic Sonatas

HRG

Whatever the range of constants we are dealing with, the probability measure on the space of all possible configuration may assign a probability of 95% to the life-friendly region. "Equal volumes" is an ill-defined concept, since we can always replace the set of constants with a new set which is in 1-1 correspondence to the old one.

Regards,
HRG.

[ February 25, 2002: Message edited by: tronvillain ]</p>

HRG

I don't think so. For simple shapes of L (parallelepipeds: each constant within a certain interval) the mapping can be written down explicitely; and IIRC those standard concepts of topology which are needed for general shapes of L are independent of the AoC.

Regards,
HRG.

tronvillain

Finally, we have the hypothesis "This die is biased towards 1063" where we already know that one of the dice is biased towards 1063. For this h, P(h)=0.5, P(~h)=0.5, P(e/h)=1, and P(e/~h)=1/10^12.

So, P(h/e)=(0.5)(1)/[(0.5)(1)+(0.5)(1/10^12)], which is approximately equal to one. We have achieved in what roll what in the previous case would have required two. It should be obvious to you by now that what knowledge you have affects what probabilities you use.

Synaesthesia

Somms,

Consider this to be the revelation of the cosmic-sperm-of-an-ant. History has now been irrevocably affected by this event. At any rate, on we go.

We, the tiny little mite sized creatures with disproportionately sized egos really have no basis by which we can judge the dispositions of the set of possible Gods. Therefore, to assign a value of 1 to P(L|G) (Probability of life, given the existence of a god) is simply to select a particular subset (or an instance in this subset) of the set of all gods. (Namely any god that creates life.)

The evidence, in this particular issue, is totally irrelevant. If this is a god-who-creates-life we are talking about, of course he’ll create life. The question is whether a priori we can conclude that an intelligent creator is more likely than any naturalistic hypothesis.

My position is that there are naturalistic processes by which life can come about and there are supernatural forces that could do the same, we just have no way a priori of determining their respective ‘likelihood’ based upon the fact that there is life.

These prerequisite attributes are simply being pulled out of the air. There is simply no grounds for assuming that an intelligent creator would require an infinite amount of knowledge, (in fact, evolution shows distinctly that a large number of truly unintelligent processes can produce life) or an infinite amount of power (You cannot imagine most power short of infinity, I strongly doubt that you or anyone can appreciate the ramifications of such finite power.), the will to do so (it’s certainly logically possible that life-friendliness could be a consequence of other goals.) or a “distinct essence”, whatever that actually means.

I will assume, Satan, that you have never been involved in the creation of a universe. I wonder how you can dictate so unreflectively on what is involved in that process.

Regards,
Synaesthesia

"As science, medicine and psychology uncover logical explanations for the unexplainable it becomes increasingly difficult to name something a miracle." -Roger Pilon, head of The International Medical Committee for the shrine at Lourdes, France.

Malaclypse the Younger

I'm not really following this--Probably because I don't know any topology.

Consider the following scenario:

I have one graph with a rectangular unit square U in 2 dimensions. Inside that unit square, I have a smaller square C which encloses an area 5% of U.

I have another graph with the same rectangular unit square U, and within it is a (slightly) smaller square C' which encloses an area 95% of U.

Are these two squares isomorphic? If so, why? If not, how does this scenario differ from the your argument?

Not trying to be diffiult, just trying to understand.