The world of quantum mechanics gives rise to both the possibility of a way to crack RSA in seconds, and the ability to create an unbreakable code.
A school experiment shows how light interacts as it passes through two slits. The light from one slit interferes with the light from the other slit, and bands appear on the wall behind. But when one single photon is fired at the same two slits, a banded pattern is still seen. This seems impossible, but quantum theory has been developed in an attempt to explain it.
Quantum theory is so badly understood, being a completely new, different, and seemingly daft idea, that there is no agreement on exactly how it works. There are two lines of thought. In one, it is considered that quantum particles (eg individual photons), can exhibit superposition, such that if there is ever a choice of what they can do, they will do both at the same time. Therefore the photon approaching the slits goes through both the left slit and the right slit simultaneously, and interacts with itself to form the banded pattern.
The other theory is just as odd. The theory of multiverse suggests that when the photon has the choice of passing through either the right or left slits, then the universe slits in two, and in one universe it goes through the right slit; in the other is goes through the left. These two universes combine again and the photons interact with each other.
Whichever of these is right, it could be possible to use them to build a quantum computer. This would be able either to do a million calculations simultaneously (superposition) or to split into a million different computers in a million different universes, and solve one calculation in each at the same time. If this could be done then the problem of factoring a huge number would take no time at all, and RSA would be totally broken.
Quantum theory can also allow new coding systems - that are so secure even a quantum computer cannot break them. They can in fact be proved to be unbreakable, something no other code has ever been, even though they are often thought to be.
The technique involves vibrating photons. Photons are made to vibrate in one of four directions: N-S; E-W, NW-SE; NE-SW. A code can be made up from these four types of particle. If anyone tries to intercept the message, then they have two choices of how to read each particle. Either they try a N-S polarised test, or a NW-SE polarised test. They have to us the N-S one if the particle is N-S or E-W (+), and they have to use the NW-SE one if the photon is NW-SE or NE-SW(×).
The way it all works is that I can send Paul a series of photons, some using the + scheme, and some using the × scheme. Paul then reads them, but does not know which polarisation to use, + or ×. So he guesses, sometimes right, sometimes wrong. After he has read the message, we speak on the phone, and I tell him whether I used + or × scheme for each photon - but since I do not say whether for a + scheme it was N-S or E-W, and for a × scheme whether it was NW-SE or NE-SW, anyone listening cannot know from the phone call what I sent. Paul and I work out which photons Paul measured with the correct orientation, and keep these, discarding the rest. This new list of correct photon measurements can be used as a key. No-one else can have found this, since anyone eavesdropping on the line would also have had to guess + or × when eavesdropping; they would certainly have got some wrong and some right, but not at the same time as Paul, so we would discard some of those that the eavesdropper got right, and keep some the eavesdropper got wrong, meaning that they wouldn't know the exact code we'd settled on.
This new secret key can be used for encryption of a message, and since it is completely random, and can change for each message, it creates totally robust codes. With nothing at all to go on, even a quantum computer cannot work out what is written. This is because if Paul's Mum intercepts the coded message:
DJFGI JFJIDJ FOI DNIOF SDNFN IDS NIFI OSDNC FIONSD FMIO NS DIONFS DNF IWE NF IONX.
The only way to work out what is written is to try every possible arrangement of substitutions (A=A, A=B, A=C, A=D...etc B=A, B=B, B=C...etc C=A C=B...etc etc). Although this would at some point generate the correct message, it would also generate every other conceivable message that had that number of letters. Consequently there is no way of working out which one of millions is correct, and the solution is impossible.
It is effectively the security of one of the centuries-old cyphers but with the fundamental improvement that the key-distribution, so often a stumbling block to code security, is completely overcome. If quantum cryptography can be implemented, then the codemakers win, checkmate.
