Schrödinger's astronaut as a quantum computer?

[Mike S.]

TL;DR Summary

Can a truly isolated human starting from a random seed deliver output like a quantum computer? If not, why not?

Consider the following scenario: a space station is put into orbit, which is absurdly well shielded from all sorts of radiation, a.k.a. "a box". You cannot make any effective observation of what the astronaut inside is up to. (This postulate may be implausible, but in the age of "weakly observed" quantities, I'm not sure it has to be) There is no communication from inside the capsule, except a single sequence of bytes sent out by laser a week after the astronaut begins computing. The astronaut is expected to solve a very difficult mathematical problem. To help him get started, he draws a random number corresponding to some literature reference to begin reading and pondering. If he comes up with no answer, he sends no output. Only if the answer is right (perhaps he even has a cryptographic method to check it) does he send a signal.

At first glance, the astronaut, like a cat, is in a superposition of states like a quantity in a quantum computer. An output he generates should be like quantum computing, so it should be possible for the superposition of states to always send output that averages out to the right answer, even though an astronaut, by subjective experience, was unlikely to have calculated it in anyone pull of the numbers.

An obvious problem: when the astronaut leaves the capsule, is he guaranteed to remember doing the calculation correctly? If so, was he ever in a superposition of states? If not, where did the answer come from?

I assume the basis of my confusion has something to do with why quantum computing involves making great effort to avoid irreversible steps. What is the equivalent of an irreversible step where a person is involved? How would you need to alter the person/macroscopic system so his actions are consistent with quantum computing that really works?

 

[PeroK]

Reading that leaves me just as confused as you!

 

[PeterDonis]

At first glance, the astronaut, like a cat, is in a superposition of states like a quantity in a quantum computer.

Why? What quantum event causes this? In the case of the cat, it's a radioactive decay which has a 50-50 chance of happening over the time scale of the experiment. I don't see any analogous quantum event in your description.

 

[CoolMint]

Summary:: Can a truly isolated human starting from a random seed deliver output like a quantum computer? If not, why not?

Consider the following scenario: a space station is put into orbit, which is absurdly well shielded from all sorts of radiation, a.k.a. "a box". You cannot make any effective observation of what the astronaut inside is up to. (This postulate may be implausible, but in the age of "weakly observed" quantities, I'm not sure it has to be) There is no communication from inside the capsule, except a single sequence of bytes sent out by laser a week after the astronaut begins computing. The astronaut is expected to solve a very difficult mathematical problem. To help him get started, he draws a random number corresponding to some literature reference to begin reading and pondering. If he comes up with no answer, he sends no output. Only if the answer is right (perhaps he even has a cryptographic method to check it) does he send a signal.

At first glance, the astronaut, like a cat, is in a superposition of states like a quantity in a quantum computer. An output he generates should be like quantum computing, so it should be possible for the superposition of states to always send output that averages out to the right answer, even though an astronaut, by subjective experience, was unlikely to have calculated it in anyone pull of the numbers.

An obvious problem: when the astronaut leaves the capsule, is he guaranteed to remember doing the calculation correctly? If so, was he ever in a superposition of states? If not, where did the answer come from?

I assume the basis of my confusion has something to do with why quantum computing involves making great effort to avoid irreversible steps. What is the equivalent of an irreversible step where a person is involved? How would you need to alter the person/macroscopic system so his actions are consistent with quantum computing that really works?

It's far from certain that very massive bodies can be put and maintained in superposition. Under any circumstance, even if completely shielded. It's usually assumed that it is possible but its not proven.

 

[PeroK]

It's far from certain that very massive bodies can be put and maintained in superposition. Under any circumstance, even if completely shielded. It's usually assumed that it is possible but its not proven.

It's impossible not to be in a superposition!

It's just not a very useful way of describing the behaviour of an object with a large number of degrees of freedom.

 

[PeterDonis]

It's impossible not to be in a superposition!

Only if quantum mechanics, as we currently understand and apply it to microscopic systems, scales up unchanged to systems of arbitrary size, and if that continues to be true once gravity is included in our quantum theory. Since nobody currently knows how to do either of those things, I would be very, very careful about making authoritative claims like this.

 

[PeterDonis]

It's just not a very useful way of describing the behaviour of an object with a large number of degrees of freedom.

It's not just that it's "not very useful"; it's that all of our experience of macroscopic systems contradicts the claim that such systems can be (let alone always are) in superpositions. Particularly when you include gravity.

 

[Mike S.]

Why? What quantum event causes this? In the case of the cat, it's a radioactive decay which has a 50-50 chance of happening over the time scale of the experiment. I don't see any analogous quantum event in your description.

Sorry, I thought that was implied when I said "random" ( I did call him "Schroedinger's astronaut" ...)

 

[PeterDonis]

I thought that was implied when I said "random"

Just saying "random" isn't enough. You can use your computer's "random number generator" to select a random number, but that doesn't create any quantum uncertainty. Or you could roll a die, or stick your finger at random into a book and pick a number of the page, etc., etc. None of those things would create any quantum uncertainty either.

 

[Mike S.]

It's not just that it's "not very useful"; it's that all of our experience of macroscopic systems contradicts the claim that such systems can be (let alone always are) in superpositions. Particularly when you include gravity.

I agree that in theory gravity should allow observation inside any space capsule. I don't know if that might end up as a "weak observation" with negligible effect or not.

I agree that we don't perceive superpositions of states. There are various well known interpretations of this, such as that consciousness causes only one option to be real, or there are multiple consciousnesses corresponding to multiple states. The main goal of a Schroedinger's astronaut experiment, if otherwise valid (which I don't think it is, but I don't really understand why) would be to determine which of those interpretations is true by experiment. If the presence of a conscious astronaut would prevent simultaneous quantum computations in a superposed state, it would certainly be interesting!

You can use your computer's "random number generator" to select a random number

I would say that you use a computer's pseudo-random number generator to generate a pseudo-random number. For cryptography it makes a difference. But I don't want to argue semantics - let's assume the astronaut is treated just like the cat, apart from being told to read different books instead of being killed by cyanide that is.

 

[PeterDonis]

I agree that in theory gravity should allow observation inside any space capsule.

Not necessarily. If the capsule is in free fall in deep space, far away from all gravitating bodies, I think it would be perfectly possible to consider it isolated from gravity for long enough to run the kind of experiment you describe.

I agree that we don't perceive superpositions of states.

That's not correct as you state it. A correct statement would be that we do not measure macroscopic objects to be in superpositions of "classical" type states (such as being in different positions).

The main goal of a Schroedinger's astronaut experiment, if otherwise valid (which I don't think it is, but I don't really understand why) would be to determine which of those interpretations is true by experiment.

You can't distinguish between any of the known interpretations of QM by experiment. They all predict the same results for all experiments.

If the presence of a conscious astronaut would prevent simultaneous quantum computations in a superposed state, it would certainly be interesting!

You wouldn't be able to tell. See below.

What is the equivalent of an irreversible step where a person is involved?

It's unavoidable. "Irreversible" means "decoherence", and the astronaut, just in order to be alive and conscious at all, has to be constantly decohering himself, because of the huge number of interactions constantly taking place between the $10^{28}$ or so atoms that make him up, the vast majority of which cannot be kept track of. So you can't use an astronaut to do quantum computing at all.

How would you need to alter the person/macroscopic system so his actions are consistent with quantum computing that really works?

You can't. See above.

 

[PeroK]

It's not just that it's "not very useful"; it's that all of our experience of macroscopic systems contradicts the claim that such systems can be (let alone always are) in superpositions.

The problem is equating macroscopic states to quantum states. Such as trying to pretend that "alive" and "dead" are just like "spin up" and "spin down". This, IMO, is the fundamental problem.

The Schroedinger's cat experiment cleverly relates the two, but in general "spin up" and "spin down" are very different properties from "alive" or "dead" or "male" or "female"!

The superpositions of a complex macroscopic object are all at the microscopic level and we don't have words to describe these microstates because generally they make no difference.

 

[Nugatory]

let's assume the astronaut is treated just like the cat, apart from being told to read different books instead of being killed by cyanide that is.

“Treated just like the cat” doesn’t mean what you’re thinking.

Schrodinger’s thought experiment is almost universally misrepresented in popular writing. His point (writing in 1935, one paragraph on the sixth page) was not that a cat might be in a superposition of dead and alive. His point was that something was wrong with the then-current understanding of quantum mechanics because it seemed to imply that the cat would be in such a superposition, a ridiculous conclusion. The problem was resolved a decade or so later with a better understanding of decoherence, which explains why cats and astronauts show classical instead of quantum behavior.

Thus, our lack of knowledge about the cat or the astronaut is not based on superposition or any other quantum mechanical weirdness. It’s just simple classical lack of information, no different from not knowing whether a tossed coin is heads or tails until we look.

 

[Mike S.]

It's unavoidable. "Irreversible" means "decoherence", and the astronaut, just in order to be alive and conscious at all, has to be constantly decohering himself, because of the huge number of interactions constantly taking place between the $10^{28}$ or so atoms that make him up, the vast majority of which cannot be kept track of. So you can't use an astronaut to do quantum computing at all.

The message I'm taking from this and Nugatory's comment below is that the reason why Schroedinger's cat and quantum computing don't agree with each other is that the cat story is simply not true. Which is illuminating, considering how often I've seen the story cited as a fundamental basis for understanding.

Let me see if I'm following this idea correctly. The cat or astronaut is in a box, and a radioactive decay occurs. The cat goes into a superposition of states, and I don't have the information to break that superposition. However, the random course of entropy means that the superposed state is going to get narrowed down into some tiny random detail, which way an electron is spinning in one of the cat's claws maybe, while every other aspect of the cat, such as whether it is alive or dead, is in fact already "solidified". When I open the box, I may find out which way that toenail electron was spinning, but otherwise nothing special happens. Specifying the toenail electron spin would the collapse the state (because now my consciousness gets split into two possibilities "just like Schroedinger's cat in the box"...) whereas specifying every other part of the cat was decoherence. In the many-worlds interpretation, the decay of the isotope created two worlds that were exactly identical apart from one piece of data that rapidly moved into the toenail electron, which is the only difference between them. How much of this (if any) is correctly said according to a modern understanding?

 

[PeroK]

The message I'm taking from this and Nugatory's comment below is that the reason why Schroedinger's cat and quantum computing don't agree with each other is that the cat story is simply not true. Which is illuminating, considering how often I've seen the story cited as a fundamental basis for understanding.

Let me see if I'm following this idea correctly. The cat or astronaut is in a box, and a radioactive decay occurs. The cat goes into a superposition of states, and I don't have the information to break that superposition. However, the random course of entropy means that the superposed state is going to get narrowed down into some tiny random detail, which way an electron is spinning in one of the cat's claws maybe, while every other aspect of the cat, such as whether it is alive or dead, is in fact already "solidified". When I open the box, I may find out which way that toenail electron was spinning, but otherwise nothing special happens. Specifying the toenail electron spin would the collapse of the state (because now my consciousness gets split into two possibilities "just like Schroedinger's cat in the box"...) whereas specifying every other part of the cat was decoherence. In the many-worlds interpretation, the decay of the isotope created two worlds that were exactly identical apart from one piece of data that rapidly moved into the toenail electron, which is the only difference between them. How much of this (if any) is correctly said according to a modern understanding?

Let's leave the cat to one side for the moment. There are QM states that determine properties like position, momentum, spin, energy etc. These apply to elementary particles and systems comprising a small number of elementary particles. The way QM works, states may be superposed, there is quantum uncertainty and dynamic properties are only meaningful as the result of measurement.

It's then a mistake to assume that the familiar properties of macroscopic objects are subject to the same uncertainties. As mentioned above, properties like dead or alive, male or female, child or adult are not simple fundamental quantum properties, but much higher level properties that are determined by huge numbers of elementary particles. They cannot be superposed, there is no uncertainty, and whether you are alive or dead, male or female, a child or adult makes sense without the act of measurement.

The puzzle, therefore, is "where does the quantum weirdness go"? Simple averaging out accounts for some of it; and, the uncertainties are on such a tiny scale that we don't notice them (large objects do, for all practical purposes, have a well-defined position and momentum and follow the laws of classical mechanics).

To rephrase the question: how do the large scale laws of classical mechanics emerge from the microscopic QM systems that make them up?

Schroedinger highlighted this puzzle with his cat experiment. Which needs an explanation. We always find a cat and an environment that are consistent with the car having died (and been dead for some time) or, having been alive all along. We find no evidence of the cat having existed in a superposed state of alive and dead states.

A further piece in the solution is the idea of decoherence, which describes how complex systems continually interacting with an external environment tend to decohere into one thing or the other. And, more technically, the complex probability amplitudes of QM are reduced to the familiar classical probabilities of classical mechanics.

Technically, when we open the box we find the cat in either a) a superposition of a huge number of states, all of which are identifiable as a live cat; or b) a superposition of a huge number of states, all of which are identifiable as a dead cat. The intermediate result, which would be a superposition of a huge number of states, half identifiable as a live cat and half identifiable as a dead cat has in fact a zero probability.

Decoherence makes this probability vanishingly small - as it does other intermediate states, where the cat is 90% alive and 10% dead etc. Statistically, these don't happen.

 

[phinds]

The message I'm taking from this and Nugatory's comment below is that the reason why Schroedinger's cat and quantum computing don't agree with each other is that the cat story is simply not true. Which is illuminating, considering how often I've seen the story cited as a fundamental basis for understanding.

Were any of these in an actual physics textbook? I'd bet they were all in pop-sci presentations which are absolutely untrustworthy.

 

[PeterDonis]

The message I'm taking from this and Nugatory's comment below is that the reason why Schroedinger's cat and quantum computing don't agree with each other is that the cat story is simply not true.

The part of the cat story that says the cat would remain in a superposition of alive and dead until the box is opened is not true, yes. The cat is constantly decohering itself, just like the astronaut, so it can't remain in such a superposition. There would be nothing impossible about setting up the scenario described in the experiment; it just wouldn't result in the cat being in a superposition of alive and dead.

Let me see if I'm following this idea correctly. The cat or astronaut is in a box, and a radioactive decay occurs. The cat goes into a superposition of states

Wrong. The "alive" and "dead" alternatives decohere almost instantly, so they can't interfere with each other. On a collapse interpretation of QM, one or the other alternative is the one that is realized and there is no superposition. On a no collapse interpretation such as the MWI, the two alternatives are both realized, but in separate branches of the overall wave function that can never interfere.

the random course of entropy means that the superposed state is going to get narrowed down into some tiny random detail, which way an electron is spinning in one of the cat's claws maybe, while every other aspect of the cat, such as whether it is alive or dead, is in fact already "solidified".

No. There is no "random detail" left. The two alternatives are entirely decohered.

When I open the box, I may find out which way that toenail electron was spinning, but otherwise nothing special happens.

No. Nothing special happens, period.

Specifying the toenail electron spin would the collapse the state (because now my consciousness gets split into two possibilities "just like Schroedinger's cat in the box"...) whereas specifying every other part of the cat was decoherence.

Wrong. Decoherence does not mean there is still some "random detail" that is not decohered. It means everything is decohered, period.

In the many-worlds interpretation, the decay of the isotope created two worlds that were exactly identical apart from one piece of data that rapidly moved into the toenail electron, which is the only difference between them.

Wrong. In one world, the cat is alive, and in the other world, the cat is dead. That is a huge difference. It certainly isn't just a difference in the spin of one electron.

How much of this (if any) is correctly said according to a modern understanding?

Practically none of it. See above.

 

[StevieTNZ]

Please remember, folks, that decoherence is a FAPP solution. It does not mean a superposition goes from AND to OR. In principle the superposition remains - it is because we cannot track the environment that we miss information to pass into the maths to conclude a superposition still exists.

 

[PeterDonis]

In principle the superposition remains

But since the branches are decohered, they can never interfere with each other, so we can never get any experimental evidence that multiple branches exist. That is why we can't experimentally distinguish the MWI from collapse interpretations.

 

[StevieTNZ]

But since the branches are decohered, they can never interfere with each other, so we can never get any experimental evidence that multiple branches exist. That is why we can't experimentally distinguish the MWI from collapse interpretations.

I'm not talking about MWI.

 

[PeterDonis]

I'm not talking about MWI.

If you are talking about "the superposition remains", you are talking about a no collapse interpretation, of which the MWI is one.

 

[StevieTNZ]

If you are talking about "the superposition remains", you are talking about a no collapse interpretation, of which the MWI is one.

Not necessarily. collapse could occur at a later time. a system + apparatus being entangled with its environment would remain in a superposition until a conscious observer comes along. That's the view I adhere to. plane @ simple

 

[PeterDonis]

Not necessarily. collapse could occur at a later time. a system + apparatus being entangled with its environment would remain in a superposition until a conscious observer comes along. That's the view I adhere to. plane @ simple

Ok, but then the superposition only remains until a conscious observer looks. But there is nothing in the math of QM that says collapse should occur then, any more than there is anything in the math of QM that says collapse should occur when decoherence occurs. So you're still saying the superposition disappears for reasons that have nothing to do with the math. Which means that this from your post...

In principle the superposition remains - it is because we cannot track the environment that we miss information to pass into the maths to conclude a superposition still exists.

...is just as applicable to your preferred interpretation as to an interpretation that says collapse occurs when decoherence occurs. In principle the superposition can remain even when a conscious observer looks: it's just that we can't track the detailed degrees of freedom inside the conscious observer's mind that would let us conclude that the superposition still exists.

 

[Lord Jestocost]

In principle the superposition can remain even when a conscious observer looks ...

What would it mean for a conscious observer, a so called "self-knowing unit", to be in a superposition. Here seems to be a contradiction with respect to the term "conscious".

 

[kalabren]

I would say that you use a computer's pseudo-random number generator to generate a pseudo-random number. For cryptography it makes a difference. But I don't want to argue semantics

Seems like you are arging semantics :-)

Personal computers have had hardware based random number generators as standard equipment since the early 2000s. Unless you have a very weird computer, there's a chip inside it that can generate a random value based on reading physical quantities such as electrical noise, instead of producing a "determininstic" value from a software algorithm.

We can debate whether this is true quantum randomness, but it's no longer correct to state that computers have only PRNG. However, it's still true that using the PRNG is of convenience to programmers as generating a lot of random values can exhaust the entropy available to the random number hardware, causing it to block or degrade in quality.

 

[Mike S.]

No. There is no "random detail" left. The two alternatives are entirely decohered.

I'm finding this one hard to swallow. A random event occurs within the "box", and I have entirely "observed" everything about it without exception (not in the sense of knowing, but in the sense of not leaving any uncertainty) before it opened, when by premise I had no way to observe anything inside? I could kind of sort of see the cat being determined to be alive or dead by some entropic process before opening the box, since those are not quantum states and there isn't a guarantee to an infinite amount of blurred information, but your statement here leaves me wondering why I don't see everything every time everywhere.

As an aside, I should say that a related thread pointed me to a paper ( https://arxiv.org/ftp/quant-ph/papers/0306/0306072.pdf ) which explains decoherence of a macroscopic system in terms of what is taken as inevitable linkage to an environment. It says the time scale of the decoherence depends only on the thermal de Broglie wavelength and the relaxation time... but I'm unclear at the moment on what the latter term really means.

 

[PeterDonis]

What would it mean for a conscious observer, a so called "self-knowing unit", to be in a superposition. Here seems to be a contradiction with respect to the term "conscious".

Without an actual physical theory of consciousness, it's impossible to answer this question.

 

[PeterDonis]

A random event occurs within the "box", and I have entirely "observed" everything about it without exception (not in the sense of knowing, but in the sense of not leaving any uncertainty) before it opened, when by premise I had no way to observe anything inside?

I didn't say you had observed anything inside before you open the box. I said the two alternatives for what you will observe when you open the box, "cat alive" or "cat dead", are entirely decohered. That means there is no quantum interference between them, so there is no way to show experimentally that both alternatives still exist. You will observe one or the other when you open the box, and that's it.

It says the time scale of the decoherence depends only on the thermal de Broglie wavelength and the relaxation time

The relaxation time is basically the time it takes for interactions with the environment to become significant. But "environment" for macroscopic objects includes all of the untrackable degrees of freedom inside the objects themselves. For example, in the case of the cat, the relaxation time is the time it takes for the quantum degree of freedom represented by the radioactive atom, which either decays or doesn't decay, to interact with everything else inside the box. Not just the cat, but the apparatus that detects the decay and releases the poison that kills the cat. There are enough degrees of freedom even inside that apparatus that decoherence will have taken place before even the cat itself is affected. In short, there is enough "environment" inside the box itself for decoherence to take place, even if the box is isolated from everything else before it is opened.