Just read this comment from https://www.youtube.com/watch?v=i2o7KbDTkgU , and it made me think of Q, Project Looking Glass, and "Future Proves Past"
Delayed Choice Quantum Eraser experiments of Scully et al published in Phys Rev Letters, Vol 84, #1, 3 January 2000. In their setup a single photon is sent toward a double slit with the BBO crystal behind the slit screen straddling the two slits. At each slit the incoming photon wave gets split into two photons, one of which is called the signal photon and the other the idler photon. The lens configuration supports capture of each of the idlers coming from the two slits. The photon paths then come to a half-silvered mirror where each photon can either pass straight through the mirror or reflect off its surface with a 50:50 chance. The data collection is done with one photon at a time. After the data are collected it is possible to deconvolute the case for which path information has been recorded from the cases for which the path information has been destroyed (erased). Looking at these separately you find that the data for path information shows the normal diffusion effect from the two slits and the erased information gives the interference pattern. What is astounding is that the choice by the half-silvered mirror, which leads to getting path information or not, occurs about 0.7 nanosecond - after the signal data are recorded! If the idler arm were made arbitrarily long (subject to the normal constraints such as maintaining coherence), you could imagine replacing the half-silvered mirror with a human and a switch. From the lab perspective, this human would be making free will decisions in our future with us observing either path or no-path decisions on bundles of photons. This is communication from the future to the present and, practically speaking, there is no hint of delay even if the human observer is far away.