5

Classical computers are usually incapable of generating true random numbers as they are based on deterministic algorithms. To overcome this challenge, one can either use pseudo-random number generators or connect to some external stochastic system (or ideally a radioactive decay).

Quantum computers seem to already solve this issue. We have already some single qubits with high fidelity. Quantum mechanics is effectively (putting interpretations issues aside) random. By applying a Hadamard transform to a set of $|0\rangle$-initialized qubits, one can measure random binary results and generate random numbers.

My issue with this challenge, it is that is seems already like a solved problem. I have seen from time to time that some researchers work on quantum random number generators (QRNG), but aside from the experimental difficulty of making a quantum computer with the right fidelity, I am unable to understand what else can be explored.

Could somebody summarise what kind of research is carried out around QRNG? Is it just about proving that the generated numbers are truly random? Or about the feasibility/realisation of such qubits? Wouldn't (in principle) a single accurate qubit suffice?

Mauricio
  • 2,426
  • 4
  • 25

1 Answers1

2

One example of a research topic beyond "hey it's random" is: how do you detect whether or not the machine is working correctly? For example, see these lecture notes on "einstein certified randomness": https://www.scottaaronson.com/qclec/15.pdf

This is more than just checking if the machine is broken. The really troublesome stuff is malicious actors trying to modify the machine to output random-looking-but-actually-predictable numbers. How do you know the machine is actually doing a Hadamard and measuring it? Maybe someone literally fried the thing with a laser!

Craig Gidney
  • 44,299
  • 1
  • 41
  • 116