What a qubit actually gives you
Quantum computers do not try every answer at once. That explanation is wrong, and the real mechanism — interference — is both stranger and more useful to understand.
The popular explanation of quantum computing is that a qubit is zero and one at the same time, so a quantum computer tries every possible answer simultaneously and picks the right one. This is wrong in a way that makes the whole field harder to think about.
It is wrong at the last step. A quantum computer can hold a combination of every possible answer. What it cannot do is read them out. Measurement collapses the state to exactly one result, and if that result were random the machine would be useless.
Amplitudes, not probabilities
A qubit's state is described by two amplitudes: complex numbers whose squared magnitudes give the probability of measuring zero or one. Being complex is the crucial part. Complex numbers have phase, and phases can cancel.
This is what separates a qubit from a biased coin. A coin has probabilities, and probabilities only ever add. Amplitudes can be positive or negative, and two paths leading to the same wrong answer can arrive out of phase and destroy each other.
Interference is the whole trick
A useful quantum algorithm is one arranged so that the amplitudes of wrong answers interfere destructively and cancel, while the amplitudes of the right answer interfere constructively and reinforce. By the time you measure, the probability mass has been concentrated where you want it.
What actually gets faster
- Factoring large integers — Shor's algorithm, exponentially faster than any known classical method. This is the one that threatens current public-key cryptography.
- Unstructured search — Grover's algorithm, a quadratic speed-up. Useful, but quadratic is a much more modest claim than the headlines suggest.
- Simulating quantum systems — chemistry, materials, catalysis. This is the application most likely to matter first, because the machine and the problem are made of the same physics.
Notice what is absent. Databases, spreadsheets, web servers, training neural networks, and routing a pothole report to the right department are not on the list and are not expected to join it.
Entanglement, briefly
Two entangled qubits cannot be described independently — the pair has a joint state that is not the product of two separate ones. Measuring one instantly determines what the other will show.
This transmits no information, which trips people up constantly. The correlation is only visible once you compare both results, and comparing them requires an ordinary classical channel. What entanglement buys is expressive power: n entangled qubits require 2ⁿ amplitudes to describe, which is precisely the scaling that makes them hard to simulate classically and useful quantum mechanically.
The reason you do not have one
Quantum states are fragile. Any interaction with the environment — heat, stray fields, vibration — destroys the phase relationships the whole scheme depends on. This is decoherence, and it happens fast.
Error correction is the answer, and it is expensive: encoding one reliable logical qubit takes many noisy physical ones, with current estimates running from hundreds to thousands depending on the hardware and the error rate. That ratio, not the raw qubit count in press releases, is the number worth watching.
What to do about it now
For almost every organisation the honest answer is: nothing, except one thing. Encrypted data being intercepted and stored today could be decrypted later by a machine that does not exist yet. If you hold data that must stay confidential for a decade or more, post-quantum cryptography is a present-tense concern rather than a future one.
Everything else is a research programme, and it will announce itself clearly when it stops being one.