Why must quantum error correction handle phase-flip errors in addition to bit-flip errors, even though classical error correction only deals with bit flips?
Think about your answer, then reveal below.
Model answer: Quantum states carry both amplitude and phase information. A classical bit can only flip (0 to 1 or vice versa), but a qubit can experience bit flips (X errors), phase flips (Z errors that map alpha|0> + beta|1> to alpha|0> - beta|1>), or both (Y = iXZ errors). Since the phase is essential for interference and thus for quantum computation, phase errors are just as damaging as bit-flip errors. The Shor code addresses both by concatenating a bit-flip code with a phase-flip code.
Classical bits have no phase, so classical error correction only needs to handle bit flips. Quantum error correction must protect the full quantum state, including relative phases. A key insight is that any single-qubit error can be decomposed into a combination of I, X, Y, Z Pauli operators. If a code can correct X and Z errors independently, it can correct arbitrary single-qubit errors — this discretization of the continuous error space is what makes QEC possible.