This post is inspired by a study of the “delayed choice quantum eraser” experiment described e.g. at
Even though the quantum eraser experiments don’t allow true “backwards causation,” this doesn’t prove that such a thing is impossible. It just proves that there is no way to do it within the commonly accepted constraints of physical law. There is at least once concrete possibility for how currently known physical law may be breakable, in a way that would allow backward causation (and, as an effective consequence, time travel – since being able to cause events in the past would mean being able to create an exact replica of oneself in the past, including a brain-state possessing the feeling of having just been quantum-magically transported into the past).
This possibility is “quantum psychokinesis” – a notion which sounds bizarre, but is apparently supported by a variety of experiments done by respected scientists at various institutions including Princeton University; see
The simplest of these experiments involve people trying to influence, by the power of concentration, random events such as the direction of an electron’s spin. A long list of experiments show that, after some training, people have a weak but real ability to do this. Over tens of thousands of trials people can make electrons spin in the direction they want to 51% of the time or so, whereas chance would dictate merely 50%. This is a small difference but over so many trials is highly statistically significant.
Hooking this kind of PK experiment up to a quantum eraser apparatus, one would obtain a practical example of reverse causation. If this kind of PK actually works, then in the context of the above “paradox” situation, for example, it really would be possible for someone on Alpha Centauri to send messages faster than light to someone back home, via biasing the direction of spin of the coupled twin particle observed on Alpha Centauri. The rate of information transmission would be extremely low, since all that PK has ever been observed to do is give a slight statistical bias to events otherwise thought random. But with an appropriate code even a very slow rate of information transmission can be made to do a lot. And hypothetically, if this sort of PK phenomenon is actually real, one has to imagine that AI’s in the future will find ways to amplify it far beyond what the human brain can do.