Talk:Elitzur–Vaidman bomb tester

Aren't the mirrors incorrectly orientated?

Yes. The mirrors are all flat. They should be curved on both sides so as to split the photon firing upon them. If the photon is split-able, then it the bomb will be diffused.

Diagram incorrect

I need to re-do the diagram because the bomb is not placed in the correct position. The following website has a better illustration: [bomb diagram], [experiment diagram] -- Thoreaulylazy (talk) 21:44, 10 October 2010 (UTC)

Glad that someone's paying attention to this article. But what point does this gedanken experiment make that is not already made by the double slit experiment, which makes the same point just as dramatically? --Vaughan Pratt (talk) 04:32, 14 October 2010 (UTC)

The double slit experiment is just as amazing, but in my limited experience, I always have to explain to people why it's amazing. It's like how, if you have explain why a joke is funny, perhaps you should tell another joke. You don't have to explain why the bomb experiment is amazing. Fool4jesus (talk) 18:01, 17 February 2012 (UTC)

I think this is a lot more important than merely being more impressive or easier to understand vs. double slit. This thing conceptually (if not practically... yet) opens the door for macroscopic exploitation of counterfactual quantum spookery in a way that the double slit experiment alone (or the cat in the box) does not. Imagine that we can achieve very long pathways (such as those that might be implemented using the scaled-down revision of the FELIX experiment.) Now, let's think of some interesting irreversible macroscopic phenomena that might take place during the delay BEFORE the photon it reaches the final detectors. The bomb tester may not be demonstrating new physics, but it *is* potentially a milestone further down the road to quantum weirdness having real technological relevancy (I mean, something beyond mere quantum computing)... and/or a real philosophical impact on even a rough and informal idea of free will.

Just imagine for a moment (and, incidentally, I once tried to write a SF story based on this premise) that the thing being measured involves a human being, with the pathway stretched long enough to accommodate human reaction times... just imagine being able to make a counterfactual measurement of a human being's free choice. Don't you think that this thought experiment would be a bit of an upgrade over the OG double slit? 184.89.44.198 (talk) 09:39, 6 November 2022 (UTC)

Yes, the bomb is at the wrong location! It should be at the lower right. See [experiment diagram], which Thoreaulylazy posted above. Holy (talk) 16:52, 10 October 2017 (UTC)

@HolyT: You are correct. I had looked at that one diagram for so long, I missed that it's on the upper path in both the original paper and the lab experiment paper as well. So, does it matter? Is it worth fixing the illustrations and the verbiage? Informata ob Iniquitatum (talk) 19:58, 10 October 2017 (UTC)

@InformationvsInjustice: @Informata ob Iniquitatum: I haven't seen the original paper and lab experiment paper. But the diagram that I linked to shows the bomb at the lower right, taking the place of the mirror (i.e., the bomb's sensor is the mirror). I think that it matters because, with the bomb in the middle of the bottom path, it absorbs the photon whether it's a dud or not. If the bomb is a dud, then it becomes a mere obstruction. Even if there is some explanation for how the photon passes through the dud bomb's sensor, the way the bomb is depicted in the existing diagram is just unnecessarily confusing. Holy (talk) 21:58, 10 October 2017 (UTC)

superposition

"... even though only one path is actually taken" That is a classical point of view and has no real meaning in quantum mechanics. The usual interpretation is that the photon does go through all possible paths, not that it somehow remotely "senses" them. Moo 02:00, 12 February 2007 (UTC)

Not when an observer is introduced. An observer, at the moment of observation, collapses the wave function, so the photon cannot be at two places at once. The bomb detector acts as the observer. —The preceding unsigned comment was added by Thoreaulylazy (talk • contribs) 21:44, 3 March 2007 (UTC).

There is always the old-fashioned point of view where the wave function "collapses" each time there is an observation. I have never liked this perspective, or even understood it, because there was never a formal criterion for when an observation is occurring. But nevertheless, this interpretation persists until today, at least for the Schroedinger equation, and it seems to be self-consistent. Therefore I took this point of view and even reinforced it in the edit I just did.

178.38.122.45 (talk) 21:26, 10 April 2015 (UTC)

Kwiat Reference?

In 1996, Kwiat et al. have devised a method, using a sequence of polarising devices, that efficiently increases the yield rate to a level arbitrarily close to one.

First of all, polarizing is spelled incorrectly. Secondly, can anyone come up with a reference to this method? I'd love to read more about it. --216.152.208.1 03:40, 16 July 2007 (UTC)

Note that a reference for this has now been added, and that "polarizing" is perfectly acceptable for US-English spellers (though I'd prefer the original "polarising") :-) 203.110.13.5 (talk) 05:58, 16 July 2008 (UTC)

One of the references is incorrect. Note 1 refers to "Paul Kwiat 1994", but none of the Kwiat entries in the bibliography have that date. --Maltelauridsbrigge (talk) 10:53, 13 November 2008 (UTC)

Im busy with the start of the new year, but this article deserves to be so much better, Ill get round to adding a quantiative analysis and a better diagram over the course of the next few months, and it will include how to increase the efficiency of the detection. —Preceding unsigned comment added by 86.141.231.205 (talk) 01:44, 3 January 2010 (UTC)

"Thought experiment"?

Given that it's actually been performed (albeit not with actual bombs) is it still fair to call it a thought experiment? Schrödinger's Cat was an untestable thought experiment designed to seem impossible and thus cast doubt on Quantum Mechanics as a whole. This experiment, on the other hand, was designed to be testable and was successfully tested within a year of its proposal--why, then, is the article titled "Elitzur-Vaidman bomb-testing problem" if we've already received a definitive answer that appears to agree with Copenhagen QM? Why is it referred to in the first sentence as a thought experiment, and only after reading two thirds of the article am I informed that it has actually been successfully tested?

I think that the successful physical experiment is a rather large point of interest--one that shouldn't be mentioned as a mere aside towards the end of the article. --Lode Runner 03:13, 3 October 2007 (UTC)

You are right--- why don't you put a statement about the experiment somewhere early? By the way, I think the reason it's still called a thought experiment is because there is reasonable consensus on what Quantum Mechanics predicts. So when someone finds a surprising prediction, people expect it to turn out like QM says, even if there is no experiment yet. The bomb testing problem is most significant not because it is experimentally verified ( everybody expected that ) but because it clarifies an important property of quantum mechanics explored in depth by later articles by Vaidman, namely that quantum mechanics allows you to experimentally test philosophical counterfactuals. This is a striking fact, which suggests a verifiable sense in which quantum mechanics is a many-worlds theory.Likebox 03:36, 3 October 2007 (UTC)

Or, to put it another way, it's remarkable in that it produced a measurable macroscopic effect that perfectly corresponds with QM's predictions (as you say, the many-worlds interpretation specifically--how else can you comprehend it, if not to say that in some "alternate universe" the bomb did in fact go off and prevent the photon from advancing any further?) This is what Schrödinger's Cat was originally designed for... but I thought that the irony of the cat was it could never be actually physically tested. It never occurred to me that someone would come up with an actual testable version of the problem. I'm not a physicist (quantum or otherwise), but I enjoy learning about it in my spare time and... I must say, I'm pretty damned shocked that we've achieved definitive experimental proof of a macroscopic "counterfactual", as you say. This is quite possibly the most astounding thing I've read in the past ten years, (and the sad part is the experiment itself is 13 years old... completed one year before I first checked out a book on Quantum Mechanics from my middle school library.)

Will do the edits tomorrow, after my head has stopped spinning quite so much. --Lode Runner 03:59, 3 October 2007 (UTC)

I had the same reaction when I heard about it. It's astonishing that you can measure a counterfactual. I was an undergrad around when the paper came out. Or rather, I should say, it didn't come out exactly as spread by word of mouth. It was supposed to be published in a big-name journal, but a big-name referee didn't feel confident to judge it's quality (I heard it from him firsthand), and another referee rejected it because "it's just quantum mechanics. Nothing new here." ! So this amazing result got published in a ridiculously obscure journal, and from there spread from graduate student to graduate student by word of mouth. We had a good time figuring out how to get close to 100% efficiency.Likebox 05:28, 3 October 2007 (UTC)

So how is this experiment any different from the one-photon case of the double-slit experiment? Is there some essential difference between the decision for the photon of which way to go after the first half-silvered mirror and which slit for the photon to go through? (When one slit is blocked, corresponding to a good bomb, the diffraction pattern disappears even with one photon.) Is the latter less amazing in some sense? Why isn't this "just quantum mechanics" as the referee said? --Vaughan Pratt (talk) 22:58, 26 September 2010 (UTC)

I took the liberty of renaming the article to reflect the experiment's status as a working apparatus (instead of an unsolvable or infeasibly-tested thought experiment "problem"), mentioned the successful test early on and added the extra category of "physics experiment." It deserves a better writeup than this--something on par with the Double-slit experiment or even the Michelson–Morley experiment, IMO, but that's a task for someone better qualified than myself. --Lode Runner 08:38, 13 October 2007 (UTC)

Just a comment: Please, if you have time, read about QM, learn what you can, and write when you feel you understand it well. If you screw something up, somebody will correct you. The sad fact of this world is that we're all stupid. If you wait for someone smart to come along, all you get is some charlatan who is pretending not to be stupid. Personally, I don't think I can give it the write up it deserves, because it's been so long since I really felt the revolutionary power of the result. I also had some ideas about it which I wouldn't be able to stop myself from mentioning and that would definitely be OR. Either Elitzur or Vaidman would probably go on and on about all the insights they've had since then, which are really great, but probably not as useful for a newcomer. So maybe you are the most qualified.Likebox 15:39, 15 October 2007 (UTC)

To me, the experiment proves the MWI of quantum mechanics. The simplest explanation is that all the bombs do actually explode - but in different future worlds. In one world around 50% of the (live) bombs explode, and 50% do not. Keeping a live bomb and knowing it is live is the equivalent of taking a peek into the box with the cat when mother nature isn't looking. It's absurd, in my mind, to imagine that the bomb was sampled without being sampled - it was sampled in another universe, it exploded in that universe, and occasionally we get to see the result of that explosion in our own universe with this elegant trick. This is how quantum computers work, and it shows that MWI is obviously true, no matter how mind-boggling that may be. Can some extension of the experiment be performed? Say, labeling all the bombs first, seeing which bomb explodes then using that information to choose which bomb is tested next? Could this allow some sort of communication between universes? --, 3 January 2008 —Preceding unsigned comment added by 76.184.217.42 (talk) 14:16, 3 January 2008 (UTC)

Quick comment--- the experiment does show that there is a cerrain many-worlds aspect to our universe, because the fact that the bomb-sensor would measure the photon can affect a branch where it doesn't happen. The wavefunction is many-worlds enough to "sniff out" that the device would measure the photon were it to go there.

But this experiment does not conclusively show that the bomb actually went on to explode in this branch. Perhaps, right after the initial measurement that branch of the wavefunction deletes itself out of existence. Using the formalism of ordinary quantum mechanics this is not possible--- there is no wavefunction reduction. But perhaps QM is not completely right. It is not conclusively established by the experiment that all stages of the measurement occur, just that the first few stages. Once the measurement cascades far enough so that a bomb actually goes off, that branch is forever sealed off from direct observation or communication. I think a careful statement is that this experiment implies, in reasonable practical terms, either Everett/many-histories/something equivalent, or a testable modification of quantum mechanics which includes a collapse mechanism.Likebox (talk) 21:57, 4 January 2008 (UTC)

Interference free quantum dot mirroring?

hi, I was thinking that to use Quantum reflection one could use Quantum Dot printed meta-materials with properties that allow quantum mirroring, to allow observation of an object without causing interference. Looks like it's not new though (is it?). I'm not a physicist, but hope that someone could make use of this. If my idea is going to the right direction. =)

Maybe I'm missing something?

It appears this test is really only determining whether or not a photon sensor works. The fact that a bomb is attached to the sensor seems to be contrived, and there is an assumption that the bomb will always explode should the sensor work. This experiment will only reveal dud photon sensors. It will not reveal bombs that are duds for any reason other than failure of the photon sensor.

You could perform this test on the photon sensors without the bombs, and obtain the same (but less mysterious) results. —Preceding unsigned comment added by Fragsworth (talk • contribs) 08:04, 2 July 2008 (UTC)

The photon sensors are entirely separate from the "bomb" apparatus. The sensors are tested beforehand and assumed to be 100% fully operational. The "dud" refers to the 'bomb' itself (yes, they use simulated bombs. Understand that the "bomb" is just a convenient, easy-to-understand example of an irreversible macroscopic phenomenon.) To wrap your head around the purpose of this experiment, you first need to wrap your head around the double-slit experiment and quantum mechanics in general. Good luck. --Lode Runner (talk) 16:42, 3 September 2008 (UTC)

No, fragsworth is correct. The experiment is merely testing if the photon sensor works. The experiment differentiates whether the wave-function is collapsed or not on the lower route. The only reason why the thought-experiment requires thinking about a "bomb" is to make us think about irreversible actions. The premise of the experiment is that successful photon measurement is synonymous with an irreversible action like a bomb explosion, and the experiment then goes on to show how a counterfactual measurement can be made under such a premise. -- Thoreaulylazy (talk) 15:35, 4 December 2008 (UTC)

I don't understand the obsession with the sensors. The experiment takes properly working photon detectors C and D for granted. If you want to test whether or not they work properly, there are much easier ways to do so than this. Observe: hook up the sensor. Shoot a photon at it. If it detects the photon, it works. This is trivial, and has nothing to do with the purpose of this experiment.

I agree that the "bomb" is just a useful and dramatic metaphor for irreversible actions... it could (and probably was) easily replaced by a simple wall that might (or might not) have a hole allowing the photon to pass through. The point is, you can detect the presence or absence of this barrier (even if it is more than just a barrier, but rather an irreversible macroscopic reaction) without actually interacting with it. Which is astounding. -Lode Runner (talk) 17:45, 22 February 2009 (UTC)

So the experiment tests whether a working sensor on the bomb absorbs the photon and therefore detonates the bomb, or a defective sensor on the bomb fails to absorb the photon, and no explosion occurs--right? That's fine, but isn't this experiment testing for information which was already fundamentally knowable? Regardless of what happens at the point of the sensor on the bomb, that sensor already existed and either worked or didn't work. It seems to me that the knowability of what happens to the photon is being confused with the knowability of the nature of the sensor on the bomb. As such, establishing which bombs are duds is not counterfactual; this is just a complicated way of uncovering information which already existed. I must be missing something here, too--? MasterAdamo (talk) 19:37, 13 January 2012 (UTC)

There is no need to preassume any kind of sensor is attached to the bomb. Assume the "bomb" is made out of a material which is inherently so sensative to photons that it will explode if it absorbs even a single one, but if the material is contaminated the photon will pass through / be reemitted unchanged. How do you test such a material? Pre-Elitzur-Vaidman bomb tester you simply could not, unless were willing to explode it. A simpler, less dramatic but still extremely impressive example has already been given: you can detect the presence or absence of a barrier without interacting with it (at least not in this universe, assuming one believes MWI) in any way. — Preceding unsigned comment added by 97.100.133.125 (talk) 22:45, 24 April 2012 (UTC)

Philosophical mumbo jumbo / "counterfactual" vs "hypothetical"

I'm cutting the mumbo jumbo about counterfactual measurements. Normally the term counterfactual measurement in QM refers to an assumed definite result measurement that could be made even though it isn't being made. In this case, a measurement not being made would correspond to an absent bomb (the measurement instrument). Here the bomb is always present and is always either 100% dud or 100% active (not a superposition of dud and active) so there is no counterfactual measurement under consideration in the usual sense of the term. Kuratowski's Ghost (talk) 23:16, 4 August 2008 (UTC)

You misunderstand--- counterfactual measurement, by your definition is a "definite result [of a] measurement that could be made even though it isn't being made". You get the answer even though the measurement has not been made.

In this case, you have a bomb, which if you measure to see whether it is a bomb or a dud, it goes off. You find out whether its a bomb or not, you get a "definite result", even though the bomb has a very low probability of going off, i.e. the measurement isn't being made.

So this is a countefactual measurement by your own definition. Vaidman says this explicitly in later papers.Likebox (talk) 17:41, 21 August 2008 (UTC)

But I see the point you are making--- this use of the word "counterfactual" is somewhat different than the use in "counterfactual definiteness", where that is used to characterize a theory that answers any question of the form "what would be the result of a measurement which I haven't performed". This problem is answering the question "What would happen if I were to shine a photon" when I haven't shined the photon. It is using the usual quantum formalism, which does not allow the answer to the question "what would be the result of a spin measurement in direction X" when direction X has not been measured. I can't think of how to state this difference in terminology clearly though. They are both counterfactuals, the spin measurement example is a counterfactual where quantum mechanics does not tell you the answer, while the Elitzur Vaidman example is a counterfactual where you can get the answer.Likebox (talk) 17:59, 21 August 2008 (UTC)

The difference is in "counterfactual" vs. "hypothetical". "Hypothetical" is an experiment you could still do but haven't yet - e.g. you know from the detector the bomb is good but you haven't set it off yet. "Counterfactual" concerns an observation you will never be able to make because you made an incompatible one instead - e.g. if you had never put the bomb through the detector in the first place, would it still have been a good bomb? I am not sure this thought experiment says anything about counterfactual (in)definiteness. -michael redman bornforthesummer@gmail.com

Plungers

I heard this as a physics folk story, and the teller used plugers. It's an illuminating picture. If you look at the secondary literature on this, I am sure you will find a use of plungers somewhere. It's not wrong, so why delete it?Likebox (talk) 19:23, 4 December 2008 (UTC)

Cut the many-world interpretation

This is not the right place to put it. The way it's positioned in the article, it suggests as if it was the interpretation of standard quantum mechanical phenomena. This interpretation is simply much more controversial than what it attempts to explain and is inappropriate in this context. JanBielawski (talk) 22:05, 19 January 2009 (UTC)

The reason someone put that in is explained in the discussion above. This type of experiment allows you to determine the answer to a simple counterfactual question "what would happen were I to shine a light on the mirror". It allows you to measure which of two possible outcomes occur without disturbing the system (it's a macroscopic version of a non-demolition experiment). You can make an operational definition of what it means to be a "many-worlds" theory: a theory is many-worlds when you can measure the outcome of at least one counterfactual, in an arbitrarily contrived setting. Then, with this definition, this experiment shows that QM is definitely a many-worlds theory.

But you could take a different operational definition for what it means to be a many-worlds theory: a theory is many-worlds when it allows you to measure the outcome of any counterfactual. In one interpretation of this stroger definition, quantum mechanics is not a many worlds theory, because of the quantum mechanical "no-cloning" theorem. You can't perform a measurement which determines the quantum state of a single copy of an arbitrary object, since any wavefunction which has a near unit overlap with that state will likely give the exact same answer for any experiment.

But you could take a third operational definition too: a theory is many-worlds when it allows you to determine (with probability approaching one) the result of any experiment on a collection of identically prepared systems without disturbing any of the systems in any way (with probability approaching one). The question is can you determine the full quantum state of a bunch of identically prepared systems in the limit of infinitely many measurements which together have negligible probability of disturbing any of the systems? I think that this is not known. So QM might or might not be many worlds in this sense, but I suspect that it is.

Another unrelated operational definition of many-worlds is when there are computations with N quantum bits which require exponential resources to reproduce classically. QM is almost certainly many-worlds in this sense. But another definition of many-worlds is when the result of any parallel search, any NP complete problem, can be computed with N quantum bits in polynomial time, and QM is believed to be not many-worlds in that sense.

The whole reason that this thought experiment is important is because it revealed an unexpected way in which quantum mechanics is operationally a many-worlds theory. Vaidman is a notable proponent of many-worlds ideas, and considers this thought experiment as a good reason to adopt a many-worlds view, so this is not a misrepresentation of the literature. Since many-worlds is only a philosophical intepretation, there is no contradiction with standard QM. You can translate anything to any of the other philosophical views, but then you end up making the whole thing look more mysterious.Likebox (talk) 22:06, 21 January 2009 (UTC)

I agree that many world theory should be cut, because it is an unwieldy, unusable philosophical conondrum. The easiest way for anybody to understand the experiment is to use their common knowledge about choosing in the context. So the photon can turn out one way or another alternatively, that's a choice. The collapse of the potentials into an actual is the choice. The point of interference is the point where alternatives come out equal. Describing in terms of choosing also fits with the propositional language of "possibility", and what "could" have happened. This is a very usable way to understand things. —Preceding unsigned comment added by 194.151.85.2 (talk) 13:58, 23 November 2009 (UTC)

sorter?

"A bomb sorter could accumulate". Hmm.... is it a sorter or a classifier? Cogiati (talk) 11:49, 27 July 2011 (UTC)

Interference vs. Interaction

In linear optics, including the experiment described here, photons do not interact with each other (they can interact with material systems, the 'bomb' being the one relevant to this paper). The proper term for the process that occurs on the second half-silvered mirror is 'Interference'. I noticed that throuout the paper, except from one instance, 'Interaction' was used instead of 'Interference'. I think this should be corrected.

Just for the record, I note that photon-photon interaction is possible in some circumstances involving highly non-linear materials or an atom strongly coupled to a photonic cavity. But, as stated above, these circumstances do not apply to the particular experiment discussed here. — Preceding unsigned comment added by 132.77.4.129 (talk) 09:57, 16 January 2012 (UTC)

Not hitting a working bomb is also a measurement

I think it should be stressed out that the presence of a working bomb is equavalent to taking a measurement of the location of the photon, even if the photon didn't hit the bomb, and went in the other direction. The cricial point is that the information about the presence of the bomb is extracted in any case, and not only when the photon hits the bomb. In the present paper this is only implicitly stated (in 'Step by step..'). — Preceding unsigned comment added by 132.77.4.129 (talk) 10:13, 16 January 2012 (UTC)

Other noticed this too, see discussion several sections further on. 178.38.122.45 (talk) 21:15, 10 April 2015 (UTC)

Newsweek link requires paid subscription

Thoroughly excised. — Preceding unsigned comment added by Rulatir (talk • contribs) 14:30, 22 October 2012 (UTC)

MWI

I removed the following, which was labelled {{Unreferenced section|date=June 2011}} {{Expert-subject|section|date=June 2011}} (so more than two years):

One conceptual way to understand this phenomenon is through the Everett many-worlds interpretation. This interpretation of quantum physics posits that whenever multiple outcomes are possible, every possible outcome in fact occurs, because the entire universe splits into multiple copies at that point, in each of which one of the possible outcomes has occurred. The parallel universes are completely independent except for the multiple copies (one in each universe) of the particle which caused the split, which can interact with each other giving rise to quantum effects.

In the many-world interpretation, the superposition behaviour of the photons in the Elitzur–Vaidman bomb tester is interpreted as the result of interactions between the parallel copies of the photon in the multiple parallel worlds corresponding to the possible states of the photon. Therefore, when a photon encounters a half-silvered mirror, two worlds are created. In one world it passes through, and in another world it reflects off the mirror. These two worlds are completely separate except for the just-split photon in superposition. The photon that passes through the mirror in one world may still interact with the photon that reflected off the mirror in the other world; specifically, their waves interfere, giving rise to the effects observed in the experiment.

See also: Many-worlds interpretation § Weak coupling

It's not just that there's no citation yet; I don't think that there's likely to be anything interesting to say. In a situation with no sentient observer, the different interpretations are generally indistinguishable (with a few exceptions). Now put Wigner's friend in there with the bomb, and you've got something interesting: not just what the MWI says, but also how other interpretations differ. (Of course, we would still need to cite a paper or something that that discusses this first.)

That said, there is an implicit claim at the very end that this is a form of weak coupling between worlds, so I left in one sentence to say that. Of course, that still needs a citation; but I reset the clock to give it a fair chance.

—Toby Bartels (talk) 01:26, 25 August 2013 (UTC)

Typo?

I can't make head nor tail of this phrase: "with photons if a usable bomb good (at least some) without detonating it." Typo? — Preceding unsigned comment added by 84.92.142.51 (talk) 16:14, 8 April 2014 (UTC)

How can the particle be absorbed by the bomb sensor and subsequently detected at D?

The article says:

"If the interferometer is aligned so the interference is constructive at C and destructive at D, then photons will only ever be detected at C. If a bomb is now placed in the lower (transmitted) path then it will block this route and so destroy the interference pattern i.e. the photon will have a 50% chance of being detected in either (but never both) detectors. Thus if a photon is detected in D there must be a live, photon-absorbing bomb. If a photon is detected at C then the bomb may be either live or dud."

I don't understand this. How can the photon make it to D if it is absorbed by the live, photon-absorbing bomb? Betaneptune (talk) 20:58, 1 February 2015 (UTC)

I agree, it can't. It also can't make it to C. So there are 3 possible outcomes at the final detectors, namely C, D, and "neither", and the "neither" result is perfectly correlated with detonation.

The bomb's photon sensor, combined with the exploding bomb, performs an observation, just like the detectors at C and D. So if there is a live bomb, there are two observation stages and two collapsings of the wave function, one by the bomb sensor and one by the final detector. (Both observation stages are visible to the experimenter.) If there is a dud, there is only one observation.

I reworded the description at numerous points to reflect this. In particular, the article now states that there is no photon detected if the bomb explodes.

If the bomb does not explode, the wave function collapses to the "upper path" eigenstate and does not split again until it encounters the second mirror.

I don't know if the original paper expresses it this way, but it doesn't make much sense otherwise.

The original text was elegantly written, but quite telegraphic and elliptic in its mode of expression. That's why I put in more detail, and in the process discovered that there are two observation stages in the case of a live bomb.

It would be possible to further improve the description by making this double-collapse point of view clear from the beginning, but that would bring us farther from the received text and I hesitate to do so.

178.38.122.45 (talk) 21:13, 10 April 2015 (UTC)

By the way, I know there are other interpretations on how superposition and measurement work other than "collapse of the wave function". Presumably the empirical outcomes of the experiment are independent of which one we take. [same poster] 178.38.79.96 (talk) 22:48, 11 April 2015 (UTC)

I think it would first be helpful to cease referring to any sort of "bomb sensor". It isn't necessary to the thought experiment and it needlessly confuses matters--as was stated above, just pretend the bomb itself is made of some exotic substance that explodes if it absorbs a single photon, but a dud bomb will be transparent to the photon. Or if you can give up the bomb fetish entirely, just say "I want to find out whether or not there's a wall in the middle of this completely dark hallway, without any photons actually traveling down the hallway. So I'll use this second hallway (which I know doesn't have a wall) and some mirrors..."

To answer your question, it seems (I apologize if I'm misunderstanding) that both of you (Betaneptune and 178.38.122.45) haven't quite grasped the implications of the double slit experiment yet. A bastardized summary might be: "there is 'what really occurs' and there is 'what could have occurred'. And what could have occurred can influence what really occurs. And that's kinda freaky." So the answer to the question "How can the photon make it to D if it is absorbed by the live, photon-absorbing bomb?" is simply that it wasn't REALLY absorbed by the photon-absorbing bomb. THE BOMB DOESN'T GO OFF IF THE PHOTON "CHOOSES" THE UPPER PATH, that's the key takeaway. But because it could not "choose" to take entire lower path all the way to the sensors (due to the presence of the live bomb blocking it), the interference pattern that would otherwise exist (in case of a dud bomb) was blocked and the photon that chose the upper path has a 50% chance of hitting D, whereas it had a 0% chance of hitting D in the case of a dud bomb (regardless of which path it "really" took) due to the interference pattern.

If the photon "really" chooses the lower path with a live bomb then no, it doesn't make it to detector D. Or C. The bomb just explodes. In this simplified form, the detector only functions properly 50% of the time (I say "properly" because the point is not to actually set off the bomb, if indeed it is a live one), although this setup can be chained together to increase reliability to an arbitrarily high percentage.

Again, the key here is that the variables affecting the thing that didn't really occur (variable: live or dud bomb. Thing that didn't "really" occur: choosing the lower path) influenced the thing that actually did occur (Thing that actually did occur: choosing the upper path. Influence: It now has a 50% chance of hitting D, whereas with a dud bomb it had a 0% chance of hitting D.) Which is very unintuitive indeed, and it's why people are throwing around terms like counterfactual definiteness and discussing the multiple world interpretation so much. 74.229.197.48 (talk) 04:41, 17 October 2015 (UTC)

External links modified

Hello fellow Wikipedians,

I have just modified one external link on Elitzur–Vaidman bomb tester. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

• Added archive https://web.archive.org/web/20071016121316/http://nonlocal.com:80/quantum-d/v2/kbowden_03-15-97.html to http://nonlocal.com/quantum-d/v2/kbowden_03-15-97.html

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at {{Sourcecheck}}).

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 18 January 2022).

• If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
• If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 17:54, 22 December 2016 (UTC)

Proposed Illustration

Hi and thanks for your contributions. Here is a proposed illustration of how the experiment works. It is a rough draft. Please give me feedback. Also, I am concerned about OR, so please consider that as well as the content. Informata ob Iniquitatum (talk) 20:22, 15 August 2017 (UTC)

Proposed rewrite

Hi all, thanks for contributing!

This topic is near and dear to me and I have never been terribly happy with the existing article. I have done a rewrite and propose it for replacing the current article. Please see it here.

I look forward to people's sundry input, feedback and revisions. Informata ob Iniquitatum (talk) 08:16, 9 September 2017 (UTC)

Hi, I just read your rewrite and like it. Only one small thing I am not so happy with: The old and your new version both say something like "the bomb both explodes and doesn't explode". I would phrase it more cautious: It is not defined if the bomb explodes or not. -Charlie- (talk) 15:10, 19 September 2017 (UTC)

C and D are switched!

They're switched in the "Results" section. Note that the diagram at the top (near the lede) has C and D in opposite positions from the series of diagrams below. Holy (talk) 00:55, 11 October 2017 (UTC)

I agree. On 24 September informationvsinjustice's addition of Figure 1 interchanged detectors C and D from their positioning relative to the main diagram at the top (near the lede). The text in the results section is consistent with Figure 1, but not with the diagram near the lede. One of the two diagrams should be fixed. (However, it would be possible to make the text consistent with both, but that would be a pity IMHO). LondonYoung (talk) 14:25, 27 October 2017 (UTC)

The question is is it worth the trouble to fix / make them consistent?Informata ob Iniquitatum (talk) 02:04, 28 October 2017 (UTC)

In my opinion, it is certainly worth modifying the text in the "results" section to be consistent with either diagram. A slightly better alternative would be to replace the diagram in the lede with Figure 1. Both of these options are pretty easy. Ideal would be to swap C and D in the top diagram, but I don't know how to edit a figure like that - so this option might not be worth it. LondonYoung (talk) 14:38, 28 October 2017 (UTC)

On the lattermost suggestion here, I have edited the top diagram to swap C and D. This alternative has the superficial advantage of consistently assigning the Constructive interference detector with C and the Destructive interference detector with D. Gizm0steve (talk) 16:09, 18 September 2019 (UTC)

Maybe we dare snip from this site for the lede illustration? https://www.st-andrews.ac.uk/physics/quvis/simulations_html5/sims/QuantumBombGame/Quantum_bomb.html LondonYoung (talk) 20:00, 28 October 2017 (UTC)

Links in notes are incorrect

The many notes that cite "Elitzur Vaidman 1993" simply link back to the Wikipedia article itself. Can someone please link them to the actual paper, if it's available, or remove the links if it's not? I will do so if I find the paper. Thank you. Holy (talk) 22:13, 14 October 2017 (UTC)

Fixed, thanks. I'm only just getting familiar with sfn. They now properly link to the sources. Informata ob Iniquitatum (talk) 17:10, 17 October 2017 (UTC)

Mechanism of the Second Half-silvered Mirror is Unclear

Why does the second half-silvered mirror always redirect waves that have undergone interference towards C? Also, that little bit of information is only contained in the figure 6 and not in the paragraph "Part 3: The second half-silvered mirror". Please add an explanation to that paragraph. LambdaKnight (talk) 15:54, 15 July 2019 (UTC)

First of all, as I understand it this picture is correct for the front-back orientations of the half mirrors (and matches the illustration at Mach–Zehnder interferometer) :

And this one is wrong:

Now to answer your question, if the bomb is a dud, the single photon takes both paths. (The same thing happens in the double split experiment.) The first half mirror changes the photon wave differently on the top and bottom paths, because reflecting off the front of the mirror is not the same as going through the the mirror. It's not just the direction that changes, but also the phase of the wave (its x-axis offset as a standard sine curve). When the waves come back together at the last mirror, their waves again are altered differently. The top path to C has reflected off the front of the first mirror and goes through the final mirror. The bottom path to C has gone through the first mirror and reflects off the front of the final mirror. The waves end up phased in sync this way and the amplitude doubles (in the sense that it was halved by considering both paths). Front reflection and going straight through glass each create their own phase shifts, but they even out here.

There are also two paths to D. The top path reflects off the front of the first mirror and reflects off the back of the final mirror. The bottom path goes through the first mirror and through the last mirror. The two waves at D end up exactly out of sync and zero out. (Like two sine waves offset halfway, which is what noise-cancelling headphones try to do.) Front reflection causes a phase shift, but back reflection itself does not, due to an optical property of mirrors: glass-on-silver acts differently than air-on-silver on light waves. But the back reflection also involves going through the glass twice, which does create a phase shift, equal to going straight through twice.

I agree the wording currently in the article is confusing and maybe wrong: "Detector D is positioned to detect the photon only in the event of destructive interference—an impossibility." The essential point is that a single photon takes both paths, and cancels itself out at D, for detection purposes. How about, "The position of Detector D is on a path where the photon's destructive interference is in effect, resulting in no detection." Or something like that.

The article on the widely used Mach–Zehnder interferometer, the machine used in this experiment, explains the optics and waves. If you look at the picture there, especially the images in the top-right corner from the detectors (our D and C), you see one image is the negative of the other. In our case, one image is all white and one is all black. And you need the weird fact (or theory) that one photon takes both paths.

Colfer2 (talk) 17:42, 17 August 2021 (UTC)

How would it be made?

Can someone please explain how the second half-silvered mirror would be made such that constructive interference would force the photon to go to sensor C, but behaving like a particle it would go to either C or D? The article seems to just say that the mirror is aligned "such that" it would behave like this, but without clearly explaining how that works, the whole thing is pretty hard to understand.

The crux of the setup seems to be "if superposition, then C, if particle, then C or D", but it's difficult to understand how that if statement could be arranged in the first place, or how that would work.50.194.115.156 (talk) 14:49, 17 September 2019 (UTC)

I am also struggling to understand this. To me, both superpositions would have a 50% chance for C and 50% chance for D. If they combine, why will they never travel to D? 82.173.133.30 (talk) 08:58, 18 September 2019 (UTC)

The idea is that two waves arrive to the second half-silvered mirror at the same time, at which point both waves split into two waves, so there will four waves total. Two of them will go towards C, and the other two will go towards D. However, the experiment is set up so that the two waves going towards C have exactly opposite phase (their amplitudes at every moment are respectively "x" and "-x") so when you sum both the result is always the constant 0 (they cancel each other out). Meanwhile, the two waves going towards D have the exact same phase (their amplitudes at every moment are "x" and "x") so when you sum both the amplitude of the original waves is doubled. 157.88.130.132 (talk) 12:12, 23 September 2019 (UTC)

SMBC comic & cleanup

There is a web comic about this (thought) experiment, and since then, this article has seen a lot of attention and changes. While the additional explanations are mostly very good, it might need a general cleanup now. --Quazgar (talk) 06:54, 20 September 2019 (UTC)

Reflected or Passed Through in Results Section?

Sorry if I am wrong, but in the Results section and the paragraph where it says 2.The photon was detected at C: the last sentence states that the photon was reflected off the last mirror, but according to the diagram to take the upper path to get to detector C the photon must have passed through the second silvered mirror. Should reflected off be changed to passed through?

Also in the next paragraph 3.The photon was detected at D: it states that the photon must have passed through the silvered mirror to reach detector D, but to get to detector D from the upper path according to the diagram, the photon must have reflected off the second silvered mirror. Should passed through be changed to reflected off?

Yes, it's incorrect, please change it. Note the text just below it: (Note: The diagram and explanation in Figure 7 unfortunately reverses the positions of detectors C and D with respect to the diagram at the top of the page. The explanation in this section refers to the initial diagram at the top of this page.). Apparently somebody fixed the figure in the meantime but forgot to fix the text. Tercer (talk) 10:23, 27 September 2021 (UTC)

Thanks, all done.49.184.88.41 (talk) 06:32, 29 September 2021 (UTC)

Upper Path not Understood

It is stated that if the Photon is detected at D then the bomb is live but unexploded. This is because the photon took the upper path. But a photon can also take the upper path when the bomb is a dude. — Preceding unsigned comment added by 213.225.35.4 (talk) 18:37, 28 October 2022 (UTC)

That is correct, but when the bomb is a dud then the photon, even if it "really" takes the upper path, also appears to take the lower path and interferes with itself. It doesn't actually take the lower path--and therefore doesn't set off the bomb--but in the case of a dud bomb it interferes with itself exactly as if it had taken both paths... and by interfering with itself, it prevents itself from ever hitting D. The double slit experiment is the most famous practical example of this phenomenon in action. Yes, this is all really weird. This is why quantum mechanics has the reputation that it has. -- 184.89.44.198 (talk) 09:16, 6 November 2022 (UTC)

Bomb always in the way, dud or working?

I don't understand this experiment. If the bomb is obstructing the lower path, won't it obstruct the photon regardless of whether it is dud or working? How can the photon keep going through the bomb if it is dud? 95.172.233.137 (talk) 21:06, 25 May 2023 (UTC)

see the section "How it works": the assumption is that the duds do not have light-sensitive sensors that absorb the photon; instead the photon passes undisturbed. ("the triggers on the dud bombs have no sensor, so any light incident on the bomb will not be absorbed and will instead pass straight through"). --Qcomp (talk) 14:55, 27 May 2023 (UTC)

Thank you, Qcomp, very much for your reply. It therefore seems to me that for the purposes of this experiment, a "dud" bomb is equivalent to a non-existent bomb. Am I correct? 95.172.233.137 (talk) 14:46, 29 May 2023 (UTC)