Live data from Hacker News

The Many-Worlds Interpretation Has Many Problems

quantamagazine.org

51–60 of 71 posts

Re: The Many-Worlds Interpretation Has Many Problems

#51

MWI seems more elegant than Copenhagen in many formulations because the problem of the relative magnitudes of the worlds is swept under the rug. If I perform a quantum coin-flip with 2/3 possibility of heads, then it creates two worlds - but I end up in one of them twice as often - why? The only solutions I've seen to preserve the relative probabilities are: 1) sufficiently unlikely worlds sometimes just disappear, o…

> If I perform a quantum coin-flip with 2/3 possibility of heads, then it creates two worlds - but I end up in one of them twice as often - why?

This question is impossible to answer until you unambiguously define "I" and "end up" first.

My answer is that there's no singular you. There's one you for each of the worlds, just as there's one coin for each of the worlds. The only difference is that the coin doesn't wonder about its "other selves", because it doesn't have the concept of self. Humans do have one, but we have no reason to believe that it is correct or meaningful in this context.

Re: The Many-Worlds Interpretation Has Many Problems

#53
post #47
post #45

Philip Ball has previously written very similar articles on his objections to Many Worlds: https://aeon.co/essays/is-the-many-worlds-hypothesis-just-a-... Whilst I share his frustration with how readily it is being taken up and the dogmatic glamour surrounding it, I feel like he consistently misses the primary issues with it. Copy-pasting a previous comment trying to describe succinctly what I see as the central prob…

> there's nothing to suggest which branch the agent will experience isn't this accurate modelling though? as far as we can tell the collapse picks a branch at random. I just don't see how this is a problem of the model.

My description there was pretty crude, apologies. The measurement problem is often loosely expressed in terms of metaphors like "live cat vs dead cat". But if you actually look at the quantum state of such a complicated system, it's this huge tangled mess of Hilbert dimensions lumped together into one big global wave function. The "alive vs dead" separation is just one possible perspective (technically it's a slicing of the Hilbert space into mutually orthogonal sub-spaces). There are infinitely many possible such ways of splitting the space up, i.e. infinitely many possible basis sets. It's not directly clear why, when we physically interact with the system, one specific basis is chosen for us to "collapse" into one branch of.

Perhaps it could be better explained as "there's nothing to suggest which branching the agent will experience splitting into".

I should note that the process of decoherence goes some way towards removing this obstacle (although not the following question of explanation of probabilities).

Re: The Many-Worlds Interpretation Has Many Problems

#54
This is a very philosophical / meta-physical discussion of quantum mechanics, and not in line with our modern understanding of the measurement process. For example, there really is no sudden wave function collapse in quantum mechanics, as every measurement process can be described as a purely quantum-mechanical process that produces an apparent "collapse" of a wave function through entanglement and decoherence. There's nothing magical or abrupt about it, today we can even perform quantum measurements that precisely control the amount of decoherence that is introduced to a system (see e.g. some of Irfan Siddiqi's seminal quantum feedback experements like https://www.nature.com/articles/nature11505, or the earlier work done in Serge Haroche's group). Taking aside the philosophical problems that people have with "pure" quantum mechanics (i.e. an interpretation that does away with the second postulate of wave function collapse) it perfectly describes our experimental results.

The largest problem with wave function collapse (which the author only briefly skims) is that you need a way to explain how it works: Basically when the wavefunction collapses we go from a fully reversible, deterministic quantum system to an irreversible, stochastic system. At which scale is this supposed to happen? Let's say one day we can build huge quantum computers (think billions or trillions of qubits) that we can perfectly control. We could then prepare a single qubit in a superposition state. We can entangle this qubit with the other qubits (the measurement system) and manipulate the state of the measurement system using a deterministic but highly chaotic control program. If we'd then measure the state of the single qubit (e.g. by preparing many identical systems and performing quantum state tomography) we'd find that the wave function has collapsed, i.e. there is no more coherence between the two qubit states. Now, after entangling the qubit and evolving the large system we could just reverse its deterministic evolution (as we have perfect control over it) and bring it back to the state it was in directly after entangling it with the qubit. If we then perform quantum state tomography of the qubit we should see that the coherence is back. Now, if we believe that wave function collapse is a phenomenon that occurs independently and not as an effect of the evolution of the large quantum system, we would expect to observe a loss of coherence in the qubit even after perfectly reversing the state of the large quantum system. The question is then of course: At which scale does this happen, and what is the physical theory that governs this behavior? Are one billion qubits enough to produce it? One quadrillion? To my knowledge, no one came up with even an idea of how to describe this. If someone finds a good theory for describing wave function collapse I'll consider it as valid theory, until then I'll stick with the "many worlds" interpretation, though I'd really prefer calling it "plain" quantum mechanics instead.

Re: The Many-Worlds Interpretation Has Many Problems

#55
post #25

Earlier quoted context omitted.

I don't think that's how the anthropic principle works, but I'm not actually sure how the anthropic principle works. To my understanding, the anthropic principle is basically: among the near-infinite set of possible physical laws, we can immediately rule out the ones that would prevent us from being around trying to figure out which physical laws govern the universe. I think there's two varieties of this. The "weak a…

If the MWI is true, then there must be some universes that would experience multiple highly improbable events. These events would be a hint to residents of that universe that the MWI is correct. Why wouldn't the usual Bayesian analysis apply here?

Imagine yourself as a resident of that "universe" (I prefer to think of it as "branch").

Why should a sequence of unlikely events be a distinguisher between different interpretations of a theory, which make the same probabilistic predictions as each other?

Re: The Many-Worlds Interpretation Has Many Problems

#56
post #8

Earlier quoted context omitted.

People ridicule the nonlocality of Pilot Wave theory. But I’d rather question locality than embrace Everett’s many-worlds theory. If you need to postulate such amazing complexity just to say basically nothing in terms of realism, then you’ve already lost. Meanwhile as one good physics professor pointed out, none of these alternatives to PWT even rule out nonlocality. So why not just assume it and be done with the wei…

The evidence against Pilot Waves keeps stacking up. https://www.quantamagazine.org/famous-experiment-dooms-pilot...

That experiment really has nothing to do with actual pilot wave theories. It just rules out a fluid-mechanical analogue of quantum mechanics (which anyone could predict anyway since fluid mechanics isn't non-local in the same way).

Re: The Many-Worlds Interpretation Has Many Problems

#57
My own problem with many worlds is (a) it seems to me to require a ridiculous amount of energy and (b) what happens when two world splits occur "at the same time" on opposite sides of our planet? Also (c), it just doesn't appeal to intuition, but it seems like much of physics does not :)

Re: The Many-Worlds Interpretation Has Many Problems

#58

Quantum suicide is a fascinating thought experiment. If consciousness is a property of a certain type of brain (or calculation?), then the concept of ‘’’me” is just a variation of the anthropic principle: I believe my subjective experience to be something unique, but there’s nothing unique or special about a human brain and the conscious experiences it produces. So even if there are billions of worlds, there aren’t b…

>One consciousness is simply extinguished forever. The other continues. If there is one instance of a widely-distributed program installed on a particular computer, and the instance of the program on that particular computer is deleted, but the program continues to exist in identical form on many other computers, then the program continues to exist in spite of the fact that one of its instances has been deleted. If c…

Great points.

I think what you and I are spending a lot of words working toward is a simple question: does subjective experience have one subject or many?

Re: The Many-Worlds Interpretation Has Many Problems

#59
post #42
post #35

Earlier quoted context omitted.

I think there is a still more straightforward and robust argument against quantum immortality: The same reasoning predicts you can never fall asleep. If futures where you are not conscious are inaccessible by the anthropic principle, then dreamless sleep cannot be in your future. Most sleep is dreamless. At 4:00am this morning I will almost certainly be in this state. But being unconscious is not an experience I can…

I don't think that's accurate. The many worlds interpretation is fundamentally an interpretation of something unobservable; if there were a way for you to observe it (by noticing you haven't been able to sleep) and 'report back' to all/most universes, it wouldn't be an interpretation anymore. (If you did a quantum suicide experiment, you might be able to 'report back' to the subset of universes where you still exist,…

Quantum immortality says something concretely different than what you're describing; it says that you always have a subjectively high probability of continued existence. By quantum immortality's reasoning, that is not the case under classical mechanics or non-Everettian quantum mechanics, where there are many situations where you have a conscious future with subjective probability of effectively zero.

Of course I disagree with quantum immortality though I agree with Everett, and I think you are kind of restating my point: Finding outselves in the purple region is not surprising and is what we should expect.

But quantum immortality as I believe it is usually stated does not require you to be ambivalent about non-existence like you say. It promises future existence, which I think is invalid.

Re: The Many-Worlds Interpretation Has Many Problems

#60

The Many-Worlds Interpretation is seriously mischaracterised in the summary at the top of the article (presumably not written by Ball himself) when it says this: " The idea that the universe splits into multiple realities with every measurement has become an increasingly popular proposed solution to the mysteries of quantum mechanics. " This is a misrepresentation of what the Many-Worlds Interpretation actually sugge…

Thankyou for clearing this up. I've been confused about this for a long time. It also explains why the decoherence theory fits in with this story.
Post reply on HN