layman/terms
Quantum words

Measurement

mezh · er · ment

In shortReading a qubit's value. The moment you look, its mix of 0 and 1 snaps to just one of them.

A qubit can be a mix of 0 and 1. But nobody ever sees the mix. The moment you measure it, you get a plain 0 or a plain 1.

Picture a dimmer dial that can sit anywhere between off and fully on. Now picture a strange rule: the moment anyone checks it, it snaps to completely off or completely on. Never in between.

That’s what measuring a qubit is like. A qubit can be in a mix of 0 and 1, but measuring it always gives a plain 0 or a plain 1. Where the dial was set only decides how likely each answer is.

Why it matters

Run it again and again

If a qubit is set halfway, measuring gives 0 about half the time and 1 about half the time. One reading tells you very little. That’s why quantum programs are run many times and the results counted.

What it isn’t

It isn’t a person looking that makes the difference. Any interaction that gets information out of the qubit counts as a measurement, even a stray bit of heat or light. That’s why quantum computers are kept so cold and well shielded.

Where the comparison breaks down

A real dimmer has a setting you could read with a careful tool. A qubit doesn’t. There’s no hidden answer waiting to be found: the mix is real until the moment it’s measured.

How the words fit together

A qubit can be in superposition. Measurement ends it and gives one answer. Entangled qubits give linked answers when you measure them.

Good to know

Because each reading gives just one answer, quantum programs are usually run thousands of times. The pattern of answers is what tells you the result.

When you'll hear it

“The state collapses when you measure it.” Looking at the qubit ends the mix and leaves one plain answer.

“We ran it a thousand shots.” The program was run a thousand times to see how often each answer came up.

“Measurement is destructive.” Once you've read a qubit, its delicate mix is gone for good.

Say it in a meeting

“How many runs does the quantum program need before we trust the answer?”