6 Comments

I don't get your response to the sampling argument. What is the sample space? I can understand what the broad possibilities for the fire squad scenario are. But, as you yourself illustrate above, the number and variety of possible universes is unfathomable.

One way to go about this discussion is to make the observation that our notion of what god is evolves the more we know. Things that were once in the deity domain (from thunderstorms to dreams) consistently move as new understanding come along. If you want to call the set of universes God, that's fine. One thing, I think, is clear. Any existing human notion of a god that has any relevance to ethics or human structures is highly unlikely. So what is the use of using this loaded term?

Expand full comment

Does it make a difference that there are very many conceivable universes? There are very many conceivable ways that the story can continue after the order to fire, but it's still clear that the true continuation, with you waking up, is unlikely. Or are you arguing that there are so many conceivable universes that we don't even know how to conceptualize probability here? I doubt that.

About the second point, the argument I made is that the universe might have been intentionally created. That doesn't map neatly to any one religion, but it does map neatly to a core part of the traditional idea of God: a powerful, intelligent Creator. So I think using the word God is fine, just as 18th century Deists did.

Expand full comment

I'll borrow an answer from Maya Bar-Hillel. What is the probability that I have in my wallet a banknote with the serial number 2051326667? Highly unlikely. But, hey, what do you know?

Of millions of executions by fire squad in your imaginary scenarios, it's not surprising that someone woke up. What you are doing is waiting until someone wakes up, look to check who it is, and wonder how come this specific guy woke up. You should ask how likely is it that some kind of life that can wonder about its own existence came to be, not how likely this specific form of life is. For that you need to know how many universes there are, what forms they may take, what is the distribution of types, and how much time (which is not even defined outside a universe) these universes have to come into existence and live out their lives. Come to that, how many multiverses there are where this can happen. Given existence, an existence that is somehow constrained in some dimension strikes me as a very special point in any topology. The most baffling thing is why existence. You may refer to God on that point, but that would definitely be a god far removed from the pantheons we have, and not have any bearings on ethics.

Expand full comment

I think that you are taking words like "wonder" as primitives. But I don't think we can really understand concepts like "wonder", "think", "interact with the environment" - concepts that make up our idea of intelligent life - without a world that is like our own in some ways. (What's an "environment" without a concept of space, for instance? What would "intelligence" mean without time?)

But, even if that is false and there are forms of intelligent life that could exist in worlds very unlike our own - universes with 9 dimensions, say, or with some weird topology - I would still argue that the set of universes which support intelligent life is very small compared to the set that doesn't.

One argument would be by analogy. Whenever we look at a reasonably constrained subset of universes, we find that the vast majority can't support complexity. Consider the set of universes similar to Conway's Game of Life (a board, a concept of "neighbour", a state of being alive or dead, and a rule for moving from time t to t+1). Conway's game generates interesting phenomena like gliders. Most similar universes don't.

I don't know how to generalize this argument to the much bigger space of possible universes - or even just possible universes with space and time - but I suspect that it does generalize.

Expand full comment

Fair enough, but that's where the sampling argument kicks in. Say there are K universes, each can support some self awareness with probability p. You argue that the probability of finding such a universe in the K-size set is infinitesimal. That sounds convincing. But the universes we can consider are, by definition, those that support considerate life. The question is, then, what is the probability of getting Heads at least once in K coin tosses, where the probability of Heads in each toss is iid p. That depends on p, which I agree is very small, but also on K, which I argue is very large.

Expand full comment

Mmm... this is the same as the anthropic principle, right? I've said why I don't think it is completely convincing. In particular, in this version, it sounds as if you think that there are many possible universes ("K coin tosses") so we only need to get lucky once. But why assume there are many coin tosses? There's only one observable universe. Assuming a multiverse doesn't seem more plausible than assuming a Creator. (It doesn't seem less plausible either - both are ways of explaining the surprise of our existence, and both seem to be untestable.)

Expand full comment