Creating perfect randomness is surprisingly difficult. Even the best modern random number generators have tiny systematic errors that make some numbers appear slightly more often than others. For everyday uses this does not matter, but in cryptography even small biases can create security weaknesses.
Researchers at ETH Zurich have now shown how to produce truly perfect randomness by using quantum physics. Their work, published in Nature, marks a major step forward. The method is called randomness amplification. It takes slightly imperfect random numbers and turns them into sequences that are perfectly random and can be proven so.
At the center of the experiment are two superconducting chips cooled to temperatures close to absolute zero. Each chip holds a qubit. A 30-meter-long cooled tube connects the chips. Microwave photons travel back and forth inside the tube and create quantum entanglement between the two qubits.
The researchers made the exact type of measurement on the qubits depend on numbers from an ordinary, imperfect random generator. They then applied a special mathematical algorithm to the measurement results. This process extracts and amplifies the randomness so that the final string of zeros and ones is perfectly random.
How the quantum setup works
The 30-meter separation ensures that, even at the speed of light, no information can travel between the qubits during the measurement. This prevents any hidden influences that could spoil the randomness. The improved Bell test used in the experiment confirms the entanglement is genuine and of high quality while producing data fast enough for practical use.
The final random numbers are certified as perfect and will stay perfectly random no matter what future analysis methods are invented. In the long term this technology could serve as a trusted source of randomness, much like atomic clocks serve as trusted sources of time. It could strengthen encryption for sensitive communications, digital identities, lotteries, blockchain systems, and future quantum-secure networks. The entire security of these systems depends on the quality of the random numbers they use.