One of the fears about quantum computers is that they will be able to decrypt encrypted communications that are currently too difficult to decrypt using classic computers. This presents a future potential security threat.
A proposed solution is Post-Quantum Cryptography (PQC), that is, cryptography that is too difficult for even quantum computers to unravel. Another element to solve the problem is Quantum Key Distribution (QKD), which you can add to PQC to make cryptography even more effective against the ability of a quantum computer to decrypt encrypted communications. QKD involves a secret key known only to the two points (transmitter and receiver). Theoretically, if someone intercepts the communication, the measurement of the key can (or should) change the state of one or more qubits, which becomes evidence that someone is trying to decrypt the communication.
What is a qubit? The question reminds me of a very old comedy routine by Bill Cosby where God is instructing Noah on how to build the Ark, only in that case it was a cubit. God is communicating with Noah and Noah is so perplexed he reaches a point where he asks, "Am I on Candid Camera?"
I'm similarly perplexed by the notion of using qubits in quantum computing. A qubit is like a classic bit, except instead of being in one of two states (0,1), a qubit can be in multiple states based on a set of probabilities. Now, the word "probability" scares me. How can quantum computing be reliable if you introduce probabilities to the state of qubits? This is, indeed, a problem in quantum computing, such that you need complicated error detection and reconstruction for each process involving qubits. When you construct a quantum computer, you have to account for noise affecting the state of qubits. So you run tests where you know the value you expect from a computation and see if the quantum computer produces the correct result. The result will almost always be different than what you expect. So, you intentionally increase the noise and measure the results. This gives you a plot of noise vs. results, which you can use to work backward to project the result you would get with zero noise.
Another method of error checking is to store many copies of a qubit, do a checksum, and then do a checksum when you read the qubits. If the checksum doesn't match, then you must figure out the probability of which qubit was the correct one.
All of this leads me to have zero confidence that quantum computing will do anything useful in the foreseeable future, let alone be useful for decrypting encrypted communications currently beyond the reach of classical computers. Time will tell, I suppose.
