Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Forrest Gump meets the Elitzur-Vaidman bomb tester

  1. Sep 3, 2008 #1

    Assuming I understand this right, we've been able to test counterfactual events (i.e. interactionless measurement) for fourteen years now. Not as a thought experiment, but as an actual functional machine. Why the hell isn't anyone doing anything interesting with this idea? Screw Schroedinger's cat; this is something that's DIRECTLY TESTABLE. Let's figure out another thought experiment here, one that's much more spooky than the poor cat.

    Suppose you built a really huge version of the interferometer. The photon's paths must be long enough for a human to react (let's say five seconds), and they must be curved. This means we're dealing with a gargantuan (and extremely expensive) interferometer, but it's one that could plausibly be built.

    The "bomb" to be tested here is a human reaction. It can be anything you like... let's start with something fairly straightforward:

    Forrest Gump is sitting on a bench holding a box of chocolates. This particular brand doesn't tell you which chocolate is which. Forrest loves the coconut ones, but he hates caramel. To avoid accidentally eating any of the dreaded caramels, he uses his vast shrimp-made fortune to build the aforementioned interferometer. He presses the start button, and waits. If the red LED lights up (which only happens if the photon takes the path which makes the "bomb" goes off) , he eats the mystery chocolate currently in his hand. IF AND ONLY IF he likes the chocolate, he presses the "tasty!" button. This button interrupts the photon from reaching its ultimate destination at the sensor (remember, this is a really huge circular interferometer. He has a few seconds before the photon will pass by the critical point.)

    Now, supposedly the method described here allows the yield rate to be increased arbitrarily close to 100%. In other words, detonations (which are undesirable) can be reduced to an arbitrarily low number. In this experiment, pressing the tasty button is the equivalent to the bomb being detonated. (You could probably design something with a "not-tasty!" button as well, but I'm trying to keep things as directly analogous as possible.)

    What does mean for our hero Gump, who's apparently made several trillion dollars and blown it all on a fabulously complex globe-encircling interferometer? Well, let's say the yield has been boosted so that 99.9999% of the time that red LED (the "eat the chocolate in your hand right now" light) never goes on. He almost never has to press the "tasty!" button. The green LED, on the other hand, will light up quite often. If it lights up, he's guaranteed to have a tasty coconut chocolate in his hand. If it fails to light up, it's either still-undetermined (50% chance of being either) or it's a nasty caramel. (There's no reason why he can't make several test runs. After a few indeterminate tests, he can always discard the chocolate as being a likely caramel.)

    Quick summary:

    Red light = must consume as an unknown and elect to press or not press the "tasty" button within a few seconds, before the photon passes the cutoff point (Red lights can be minimized to some arbitrarily small probability.)

    Green light = confirmed tasty, safe to consume at your leisure. (As red lights are reduced in frequency, green lights remain unaffected.)

    No light = indeterminate chocolate

    Here's the kicker... here's where things definitely get spooky:

    1. The possibility of the red LED lighting up is very slight (if we use the above percentage, it's one in a million.) He could eat a box of chocolates every day for the rest of his life and though he'll see many thousands of green lights, he'll likely never see a single, dreaded red light.

    2. But his ATTITUDE about the possibility of a red light seems critical, right from the very first test run. If he's secretly planning to cheat... if he's really going to throw away the chocolate if ever that red LED lights up, then the chocolate-tester won't work... the green LED will never, ever light up.

    This machine is amazing enough because it predicts what's in the center of a mystery chocolate, but it's even more amazing because, by virtue of its effectiveness, it predicts the actions of a human being. Via many-worlds, it's saying that in an alternate universe, the red light DID go on and Forrest DID eat the chocolate without knowing either way.

    That's what's so crazy about quantum computers... under the many-worlds interpretation, they're basically using the computing power of the computers in alternate universes to solve problems in THIS universe. And the chocolate tester, by virtue of its effectiveness or lack thereof, is telling us what Forrest Gump is doing in an alternate universe (in which the red light goes on.)

    If someone cares to explain what happens (either in the bomb tester or Gump's chocolate tester) under other interpretations of QM, I'm all ears. I'm especially curious as to what happens under other interpretations if Gump goes in with the intent to cheat (throw away the chocolate and take no other action in event of a red light.) I'd also be interesting in knowing if my extension of the Elitzur-Vaidman bomb tester experiment is flawed in any way. (I have no formal physics education.)

    I have some more thought experiments (including some very nasty ones I've dubbed "Schroedinger's slave"), but I think that this example nicely captures the essence of what I'm driving at here.

    Let me just say this: quantum computing pales in comparison to a device that will let you stare into the heart of a man.
    Last edited: Sep 3, 2008
  2. jcsd
  3. Sep 9, 2008 #2

    Come on, guys. The Elitzur-Vaidman bomb tester is the most startling experiment of the last 20 years, and yet there are less than a dozen posts here that mention it (and none that focus on it specifically.)

    It essentially letting us peek into Schrodinger's box. It allows us to measure what the results are IF event X happens, even if event X doesn't happen in this universe. The implications are staggering.

    Maybe my box of chocolates thought experiment was too lame... Here's another example:

    Let's say a thousand years from now someone invents a time machine, incredibly powerful planet-sized particle accelerator, or some other megalithic machine that has the capability to do tremendous things, but also could theoretically destroy the universe (e.g. creation of new forms of non-baryonic matter, destruction of a false vacuum, etc.) What do you do?

    With an Elitzur-Vaidman bomb tester, you can find out if the universe will be destroyed without actually flicking the switch.

    This device, which was built and successfully tested in 1994, LETS YOU SEE INTO ALTERNATE (I.E. COUNTERFACTUAL) HISTORIES.

  4. Oct 7, 2010 #3
    I don't know whether it is against the rules to reply to a two year old post, but I felt this needed an answer.

    The problem with this proposal is that Forest "observes" which arm of the interferometer the photon passed through and as a result interference has been destroyed. The system has decohered. Or, in the language of Everett's many worlds, there will be a Forest who saw the red LED light up, tasted the caramel, and didn't like it so didn't press the "tasty" button but rather let the photon continue on its path through the interferometer. However, the "world" that that Forest lives in is different than the "world" of the Forest who never saw the red LED light up. You want the photon to pass through both arms of the interferometer and come back together to observe the interference, but this doesn't happen. The which-path information of the photon is entangled with Forest and so the two paths can't cross again.

    To summarize, the Elitzur-Vaidman bomb tester only works if, in the case of the "dead" bomb that doesn't absorb the photon, the bomb doesn't keep any sort of record of the photon having passed through it. I suspect that the situation is the same with the improved version of the experiment by the Kwiat group that you link to.
  5. Oct 8, 2010 #4


    User Avatar
    Science Advisor
    Gold Member

    Welcome to PhysicsForums, dstahlke!

    It is fine to reply to any thread. There have in fact been a number of tests that seek to look into this area. Hardy's Paradox is closely related, for example, and has been examined by Steinberg and Lundeen.
  6. Oct 8, 2010 #5


    User Avatar

    Staff: Mentor

    No, it isn't against the rules. It simply looks silly if you don't realize the post is old and you look like you're trying to engage in a conversation with someone who may not even have visited this forum in two years. :smile:

    But you made it clear that you know this is an old thread, and you've contributed something intelligent to it for the future record. So, welcome to PF!
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?