A classical computer bit is either 0 or 1. A quantum bit — a qubit — can be 0, 1, or any superposition of both at the same time. With multiple qubits, this stacking gets exponentially weirder. Two qubits can hold four superposed states simultaneously. Ten qubits can hold 1,024. Three hundred qubits can hold more states than there are atoms in the observable universe — all at once, all in a single device the size of a wine bottle.

This is not a metaphor. The qubits genuinely occupy all those states simultaneously, until measured.

The trick of quantum computing is to design an algorithm where the right answer's amplitude grows while wrong answers cancel out. When you finally measure the qubits, the right answer pops out with high probability. Peter Shor proved in 1994 that this approach can factor large numbers exponentially faster than any classical computer. Most modern internet encryption depends on factoring being slow — so a working quantum computer of sufficient size would, in theory, break it.

The catch: qubits are extraordinarily fragile. Touching air, light, or any vibration destroys their quantum state — a process called decoherence. Today's best quantum computers are kept in isolation tanks, cooled to milliKelvin temperatures (colder than deep space), and shielded from electromagnetic noise. A single qubit error rate of even 0.001 means a 100-qubit calculation typically fails.

In 2019, Google reported its 53-qubit Sycamore had achieved 'quantum supremacy' — solving a contrived problem in 200 seconds that would take a classical supercomputer 10,000 years. The computation was useless. The point was that it was possible at all.

Useful quantum computers may be a decade away. Or never.