Quantum Bits & Non-Binary Math

A gentle explainer for someone who already thinks in models, probabilities, and careful language — you.

A classical bit is a coin on the table. A qubit is a coin still spinning — until you look.

Binary is a choice, not a law of nature

Digital computers made “0 or 1” feel inevitable. Actuarial work already knows better: severity is continuous, frequency is discrete but not binary, and risk scores live on spectra. Quantum information formalizes a similar intuition: before measurement, a qubit’s state is a unit vector in a two-dimensional complex space, not a forced yes/no.

The state of one qubit

We write a general pure state as

\[ |\psi\rangle = \alpha |0\rangle + \beta |1\rangle, \quad \alpha, \beta \in \mathbb{C}, \quad |\alpha|^2 + |\beta|^2 = 1. \]

The complex amplitudes \(\alpha\) and \(\beta\) are not probabilities themselves — their squared magnitudes are. That distinction matters. It is why interference can cancel paths (destructive) or reinforce them (constructive), something classical probability mixtures cannot do.

Bloch sphere: the “non-binary” picture

Up to a global phase you cannot observe, a single qubit state can be drawn as a point on a sphere. The north pole is \(|0\rangle\), the south pole is \(|1\rangle\), and every other point is a superposition with a relative phase. Measurement along a chosen axis is like asking a directional question — and the answer collapses the beautiful continuum back to a classical bit for that observable.

Measurement is a model, not a vibe

When we measure in the computational basis, we observe 0 with probability \(|\alpha|^2\) and 1 with probability \(|\beta|^2\). Afterward, the state is whatever we observed. That irreversible update is closer to Bayesian conditioning than to “deleting information by magic” — except the underlying amplitudes carried phase relationships classical posteriors do not.

Why language & predictive people should care

• Representation. Embeddings already live in high-dimensional spaces; qubits are a tiny, highly structured cousin.
• Uncertainty literacy. Distinguishing amplitude, probability, and belief keeps you honest in model talk.
• Cryptography timelines. Understanding qubits demystifies which algorithms are threatened and which are not.
• Shared vocabulary. This page bridges Language (local LLM chat) and Quantum (interactive notebook).

A tiny worked intuition

Equal superposition:

\[ |+\rangle = \frac{1}{\sqrt{2}}(|0\rangle + |1\rangle) \]

Measure once: fair coin. Apply a Hadamard again before measuring and — depending on the circuit — you can recover certainty. That “undo” is interference, not luck. Classical randomization cannot rewind itself that way.