Does God play dice with the universe?

[jack]

It only recently occurred to me that people who objected to the non-deterministic nature of quantum mechanics were *right* if you think many-worlds interpretation is correct.

It's like, since Newton, physicists had a tacit assumption:

#1. The laws of physics are deterministic.

But then we discovered lots of evidence for QM that we couldn't ignore, and many people adopted a different assumption:

#2. The evidence for QM

And these seemed contradictory, so people who had #1 were (rightly) suspicious of #2, but people who accepted #2 felt they had no choice but to reject #1.

But was also assuming without even realising:

#3: there is only one universe, not a giant number of parallel universes

And it turns out that if you drop #3, you can keep #1 and #2.

Now, I don't think that's sufficient reason by itself to assume MWI. There are lots of other paradoxes that disappear (if QM works the way we think it does, though many physicists still think that is premature). But it's interesting that we might have to drop #1 one day, but not yet.

And I knew all that _in theory_, but I'd not actually stopped to think about Einstein's "god plays dice" quip since I learned slightly more about MWI.

Comments

[jack]

a "causal DAG"

I like this description.

Although IIRC the equations work just as well whether there are cycles or not, as long as it's locally directed, we just don't know if that has actually happened, or what it would be like to live in...

[ptc24]

CTCs: I think part of the reason I want a "causal sequence/DAG" is that closed timelike curves lead you to think "why that curve and not some other?". The equations can work, but with CTCs you potentially - or possibly even necessarily - have a bad case of the creation-of-information paradox.

I think my idea of a "causal sequence" is overloaded. One use for it is in the "entropy's arrow" thought experiment; imagine a low-entropy state at some point in time, and higher-entropy states either side. You could make a case for the causal sequence to be radiating outward from that low entropy point in both time directions. However ISTR that at least some ideas of the Big Bang[1] have the idea of a point in curved spacetime where all directions are into the future, a sort of North Pole of time... but then you could use the "causal sequence" gambit to ask what was causally prior to that.

As I say, it's not an idea I'm committed to, it's one I need to... stop thinking about and start discussing.

[1] Someone suggested a better term would be Everywhere stretch - warning! YouTube!

[jack]

As I say, it's not an idea I'm committed to, it's one I need to... stop thinking about and start discussing.

Yeah, but I agree it seems promising: I don't know exactly how the different sorts of casual sequence relate, my idea is slightly different, but it sounds right that they're related but different...

closed timelike curves lead you to think "why that curve and not some other?"

Yes... this is indeed the problem. But I'm not sure if it's a worse problem than "why the universe is the way it is" anyway, or just one we're less used to discussing. After all, we don't have a good answer for "why there should be one contiguous timeline with no loops" other than "that's what we're used to".

(If either case is 'caused' by something in which the universe's timeline is embedded, that could obviously determine either, but we have no good way of knowing anything about it)

you could use the "causal sequence" gambit to ask what was causally prior to that.

Yeah, that's a good description. People's intuition says there must be something temporally prior to the big bang, and it seems we actually have no good evidence for that (or even any idea what it would mean). But they're right that something must be casually prior, in your description.