Notes
The notes are the building blocks of this notebook: each one takes a single concept, claim, or investor question and works through it from the science up. They are organized below by topic, with the newest notes pinned on top. Explore how the ideas connect on the Research Board.
The best-known quantum optimizer writes every candidate answer’s score into phase, then lets a mixer make those phases interfere so better answers show up more often. MaxCut as the worked example, the cost and mixer layers as an actual circuit, what extra depth buys and costs on real hardware, and why a clear mechanism is not yet a proven advantage.
Read the note →One prize, a million locked boxes, no clues: classically about a million tests, with Grover about a thousand. How an oracle’s sign flip and a reflection about the average turn phase into probability, why about 785 iterations is optimal and more can hurt, why the speedup is only quadratic, and the reusable idea behind it, amplitude amplification.
Read the note →An answer can sit in the phases of a quantum state where no measurement can reach it. The QFT rearranges the state so those phases interfere into a bit string you can read: the transform written out, the arrows that cancel or line up, the exact case inside phase estimation and Shor’s algorithm, and why this is not an exponentially faster FFT for ordinary signal processing.
Read the note →Both answers at once, and hard to hold: the basic unit of quantum information.
The particle that interferes with itself, and why that is a resource.
The price of looking: reading a qubit ends its quantum behavior.
The question a measurement asks, and how it differs from an outcome and an average.
The bond Einstein bet against, with correlations nothing classical can match.
Two qubits, one shared state: why matching answers alone prove nothing.
A solved idea, an unsolved machine: how qubits are steered.
A recipe written in gates, from input to answer.
Turning phase into a number: the interference step inside phase estimation and Shor’s algorithm.
Hiding a molecule’s energy in a phase, then reading it back as bits.
Finding one marked item among N with about √N checks, and why that is the limit.
Turning a score into a probability: the hybrid optimizer, and what its depth buys.
All together, or not at all: one coherent state shared by the whole register.
How to catch a computer being quantum, with a score no classical trick can beat.
The qubit families compared, and why this is a winner-takes-most bet on physics.
The fragile raw material every quantum machine is built from.
The clock every qubit races before its information fades.
The cost of distance: which qubits can talk to which.
Why a good qubit will not stay good: noise that jumps between two levels.
A bit hidden where noise cannot reach, and still contested.
A test built to not fool itself, gatekeeping Microsoft’s qubit claim.
Fixing what you cannot look at.
Safety in numbers: many fragile qubits acting as one dependable one.
Computing right through the noise.
Why a useful machine needs thousands of physical qubits per logical one.
The probability an operation did what it claimed.
How an investor should read a headline full of nines.
The scoreboard that isn’t.
