Information transfer using entanglement?

  • #26
631
132
I see your point: in your case, if B would be -1, A cannot other than be -1 too, so the outcome of A would have been unchanged. And how can something that did not change have an effect, right? But I'm not speaking of outcomes of A, but of settings of A. Given a different setting at A, would that allow for a different outcome of B?
You can not try different settings at A in a single given experiment. Whatever the setting at A, B will get 1 or -1 with probability 1/2.
I suspect that the above sentence does not satisfy your concern, but I still don't know why.
 
  • #27
entropy1
Gold Member
887
53
You can not try different settings at A in a single given experiment. Whatever the setting at A, B will get 1 or -1 with probability 1/2.
I suspect that the above sentence does not satisfy your concern, but I still don't know why.
There is a caveat to go with that, I suspect: A's and B's outcomes are not independent. They produce a correlation (generally). However, both A and B produce a practically random pattern that we can't influence, so we can't manipulate their outcomes. What we can manipulate however, is the settings (orientations of apparata). My proposal was that, so that de correlation stays satisfied, we have different random patterns at different settings. So my question occurred: do the random patterns change, and which of them? At different settings, either A has to adapt do B, B to A, or something else I don't get. In any case the local outcomes are surely one thing: random. So we can't directly observe a possible change in the patterns. My point was that something in the patterns has to adapt to the correlation, and why not B, when changing A? So I'm not saying that this is the case, but that the information we can extrude of the experiment does not prohibit it!

I am not defending information transfer using entanglement, but only because the experiment stays silent on it. So that leaves the possibility of 'influence'.
 
Last edited:
  • #28
morrobay
Gold Member
711
112
Given a different setting at A, would that allow for a different outcome of B?
The tables from this paper address your question: Table 1 shows perfect correlations for twin state photons with parallel polarizers at detectors A and B both at 00
Table 2 shows data from detector A at again 00 and detector B now at 300 and correlations in accord with cos ( β-α)2 = .75
Note that with setting and outcome changes at B , data is un affected at A . Also note this is similar to my post # 15 above,
http://image.ibb.co/n1B4cQ/img009.jpg

Click on above then click again on image to enlarge
The tables are taken from this paper pages 7 & 8
https://arxiv.org/pdf/0902.3827.pdf
Also at: Am. J. Phys, Vol. 78, No 1. January 2010
 
Last edited:
  • #29
entropy1
Gold Member
887
53
Thanks, @morrobay, I will read it later.

I thought of the following: Suppose you have filter A at 90°, and filter B at 0°. Suppose you measure a singlet pair '-1' at A and '1' at B.

Now, if we would have measured both A and B at 0°, either both would have to measure -1, or both would have to measure 1.

This means that either A or B would have to change sign. Now why would there be a preference for A or for B? So, B could have changed, however, it could have been due to random chance!
 
  • #30
entropy1
Gold Member
887
53
If we assume outcomes of A and B are resembling randomness, then, if we pick the angle between A and B so that there is no correlation, one could observe:
  1. All B's are randomly distributed over A='0',
  2. All B's are randomly distributed over A='1'.
However, if we pick the angle so that there is 'more correlation', then:
  1. There are relatively more B='0' than B='1' given A='0',
  2. There are relatively more B='1' than B='0' given A='1'.
Similarly for anti-correlation and also for swapped A and B.

If I measure A='1', then in the second scheme there is more likelyhood for B to measure '1' than to measure '0'. So in the subset of A='1', B is not purely random anymore. However, we can't influence the outcome of A, so we see randomness at B. But could we say there is a pressure on B to consort with outcomes of A, in which we only have a say by increasing or decreasing the pressure by changing the angle?
 
Last edited:
  • #31
entropy1
Gold Member
887
53
@morrobay: thanks for your example in post #15 and the example in the paper. I am sorry, with all due respect, but I am not sure what you are saying here. Are you saying that the outcomes are interdependent, and that they are not dependent on the settings?

Then I would agree, with the caveat that the outcomes may depend on the settings. :smile:

I think that there is - in principle - not enough information to establish how the outcomes are created. The situation is symmetrical, so it could go either way. B depends on A and vice versa. However, that is my point: because there is not enough information in principle, that leaves room for the possibility that B (outcomes) depends on A (outcomes/settings). However, because the opposite is also possible, it seems difficult if not impossible to substantiate a claim of causality or even influence.

That said, I think the outcomes depend on the settings, because the correlation depends on the settings. So if one were to rule out 'an effect' (non-locality) one would have to stick with the outcomes depending on the local settings (like you seem to do?). One step further is to suggest that the outcomes might interdepend non-locally (of 'the other' settings), like I do. :smile:
 
Last edited:

Related Threads for: Information transfer using entanglement?

Replies
10
Views
359
Replies
12
Views
8K
Replies
1
Views
2K
Replies
2
Views
4K
  • Last Post
2
Replies
43
Views
3K
  • Last Post
2
Replies
29
Views
4K
Replies
6
Views
2K
Replies
4
Views
1K
Top