Hi everyone this is my first post so please be nice :)

Okay so imagine this setup:

Let's say that we do the classic double-slit experiment with a twist.

We all know how the double slit experiment works (if not please look it up)

If we leave the detector in place, but "turn it off", then it goes back to an interference pattern (indicative of a wave) because there was no measurement observed.

And every test of Wheeler seems to validate it.

If you don't know about Wheeler, it goes like this:

(This is over-simplified) Okay so imagine there's a mirror on the moon, or mars, or wherever, it just has to be far away enough to have enough time to do this switch-o-change-o move:

Shine a light (a laser really, it's more complicated than that but stay with me...) on a mirror here on earth, that will hit that far away mirror, and be reflected back to us and hit a detector.

In fact here's an even simpler analogy:

Imagine the same setup, but it's a TV remote. And you want to change the channel. So the (infrared) light first hits a mirror here on earth, then hits the far away mirror, then hit's your TV's remote sensor.

But after the light has bounced off the first mirror, and before the light hit's the TV, we move that first mirror out of the way.

The TV channel should still change, but it doesn't.

This is Wheeler's delayed-choice experiment over-simplified.


The double slit and Wheeler's experiment have been tested and verified.

Strange but true.

So what if we do this...

Let's mix the two experiments and store the information on a hard drive.

We could call making the observation a "1" And not making the observation a "0"

We could order (keep track of) each of the observations.

We could do this thousands of times.

We could send information as simple as what are the winning lottery numbers.

We could use error detection and correction and CRC checksums just like computer networks already do. We could send the same message 100 times, just to make sure we got the right message.

In theory, information cannot be destroyed, but we can make it effectively un-readable, in other words, to ensure a "0", just permanently delete that part of the hard drive.

And in fact, once the message has been verified that it has been sent to the past, we could destroy the hard drive altogether, either by drilling a hole in it, which is very effective, or by melting it down.

If we really wanted to make the information on the hard drive unavailable to us, we could launch it into a black hole.

Anyway, you get my point I hope.







  • 1
    $\begingroup$ Hi Brandon, this is just a general comment for any future posts. Brevity is appreciated here. The users in a position to answer your question (s) will be familiar with your references, a link to an appropriate Wikipedia article is enough. So you could really start your question from So what we could.. Using uppercase is not needed, it seems like shouting. Why I am saying this is because a good question can often be overlooked if you have already lost the reader by too much wordage at the start. Anyway, best of luck with it :) $\endgroup$
    – user179430
    Dec 26, 2017 at 21:28

1 Answer 1


Here is a youtube video that gives a better explanation of your idea (which is an extremely common misconception): Quantum Eraser Lottery Challenge | Space Time | PBS Digital Studios. And here is the followup video where they explain why it doesn't work: [...] + Quantum Eraser Answer | Space Time | PBS Digital Studio.

If we leave the detector in place, but "turn it off", then [the pattern on the screen] goes back to an interference pattern (indicative of a wave) because there was no measurement observed.

And every test of Wheeler seems to validate it.

You're mixing up two different experiments. In the delayed choice experiment, you're not deciding whether or not to have a detector present. You're deciding whether or not to insert a beam splitter or not.

Also, in the delayed choice experiment, the interference pattern doesn't literally appear and disappear on the screen depending on your choice. Your choice is to measure left vs right or else use the beam splitter to measure (left+right) vs (left-right). Then you group each screen hit based on the corresponding measurement result. Then within the groups you find wavy distributions (or not).

If you don't have the left vs right or the (left+right) vs (left-right) measurement result, you can't do the grouping. If you can't do the grouping, you can't do the wavy-vs-flat test. Which kills the whole idea.

  • $\begingroup$ Ok so I'm actually re-arranging the thought experiment. And I can in fact do the grouping, we just make groups of say 100,000 "hits" $\endgroup$
    – Brandon P.
    Dec 28, 2017 at 8:46
  • $\begingroup$ And I really appreciate those videos. Let's modify the setup. No entanglement. I love entanglement, it's heart and soul of whatever quantum mechanics really is, or at least seems to be. But entanglement always seems to be the great no-you-cant-do-this. And that makes intuitive sense to me. So anyway let's takes entanglement out of the experiment altogether. I'm saying let's do the double slit experiment, with and without our future knowledge of the result $\endgroup$
    – Brandon P.
    Dec 28, 2017 at 8:52
  • $\begingroup$ Just store the results of the experiment on a hard drive, or any media, and either just don't look at the results in the future to represent a zero, or if needed delete that hard drive, or destroy it. Otherwise observe it to represent a one. $\endgroup$
    – Brandon P.
    Dec 28, 2017 at 8:56
  • $\begingroup$ @BrandonP. If you do that you'll find that there's never an interference pattern. The erasure step comes down to a choice of what to measure, not whether or not to destroy a hard drive. Destroying a hard drive will not cause any interference pattern to spring into existence. $\endgroup$ Dec 28, 2017 at 20:35
  • $\begingroup$ Thank you Craig for your response. I found this today on youtube. This will hopefully clarify things. Please watch this: youtube.com/watch?v=xo176uIPmbY This is exactly what I am asking about. I only want to change two things in the thought experiment. Instead of 102, just make it 12. And instead of randomly selecting 50% of the slit data to destroy, do it selectively, on purpose. Just imagine being able to send the results of a 3-ball lottery. Any number between 000-999. 10 bits is all you need for any number between 000 and 1,023. I feel it wouldnt work. But why not? $\endgroup$
    – Brandon P.
    Dec 29, 2017 at 8:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.