Counterfactual what?

If quantum cryptography and quantum computing seem to defy any reasonable intuition, counterfactual computing pushes this to an extreme limit. In counterfactual computing, we use the fact that a quantum system is in multiple states at once, but use this in an unusual way.

Instead of having a quantum system that is in multiple state that correspond to different calculations, like we do with quantum computing, in counterfactual computing, we think of a computer as being in two states at once: on and off. So if we interact with a computer, we will be interacting with it as it were both on and off at the same time.

Carrying this to the next logical step, we can then interact with a computer that is turned off, but get information out of it as if it were actually on. This may seem to make absolutely no sense, but it has actually been demonstrated in a laboratory by researchers.

Fortunately, there seem to be limits to what this technology is capable of, and it may turn out to be impossible to take advantage of its bizarre properties. Otherwise we could imagine assuming that a computer is in states that correspond to either having finished cracking a 256-bit cryptographic key or not having finished this calculation, and be able to break military-strength cryptography just by interacting with a computer that is in these two states. This would allow such attacks without even having to spend any computing resources on the attack.

With any luck, researchers will find that counterfactual computing is little more than a laboratory curiosity that only works in small, well-defined experiments, and is not suitable for use in anything more than trivial demonstrations.