Here’s a summary of what I’ve learned so far.
1 of 2 Quantum computing is not faster classical computing. It’s a fundamentally different relationship with reality.
A classical computer knows things. A quantum computer holds possibilities in tension until it’s forced to choose.
2 of 2 AI is software that teaches computers to learn, make decisions, and solve problems by recognizing patterns in huge amounts of data. Quantum computing is hardware—a fundamentally different way of building a computer that uses quantum physics to process information.
Through these two points of understanding, I’m building an understanding of possibilities relationship between AI and quantum computing. So. I’ll end the summary of my learning progression effort here and proceed to learning about the next list item in the progression.
This is tricky, because doing so suggests I’ve “mastered” an understanding of the AI + quantum computing conceptual skeleton—which I haven’t.
However, I want to go one step beyond so that I can understand what it is I don’t understand.
One step beyond is linear algebra.
I asked Claude to explain linear algebra:
Imagine you have a list of numbers. That list is a vector. In quantum computing, every quantum state — like whether a qubit is 0, 1, or somewhere in between — is described by a vector. It’s just a precise way of saying “here’s exactly what this quantum thing looks like right now.”
Now imagine you have a grid of numbers called a matrix. When you apply a quantum operation — like flipping a qubit or putting it into superposition — you’re multiplying that matrix by your state vector. The math spits out a new vector, which is your new quantum state. That’s literally how quantum gates work.
Tensor products are how you combine two qubits into one system. If qubit A has its own vector and qubit B has its own vector, their combined state is built using the tensor product. Two qubits become a four-number vector. Three qubits become an eight-number vector. It grows fast.
Here’s the inversion: without this math, you’re just waving your hands. Words like “superposition” and “entanglement” sound cool but explain nothing. With the math, you can predict, design, and compute. The language makes it real.
AND just like that, the holes in my understanding are exposed.
For example, I lack sufficient understanding on state vectors:
A matrix is just a rectangle of numbers arranged in rows and columns — like a spreadsheet. Think of it as a set of instructions frozen into number form.
When you multiply a matrix by a vector, each number in your new vector is calculated from a specific combination of the original vector’s numbers, weighted by the matrix’s values. The matrix transforms the vector — rotating it, stretching it, or flipping it in mathematical space.
In quantum computing, each quantum gate is a matrix. Applying a gate to a qubit means doing exactly this multiplication. The math determines the outcome precisely — no guessing required.
I also lack sufficient understanding on quantum gates:
A quantum gate is a basic building block of a quantum computer — similar to how AND, OR, and NOT gates are the building blocks of regular computers.
Their purpose is to manipulate qubits. Without gates, a qubit just sits there. Gates are what actually do something — they can flip a qubit, put it into superposition, or create entanglement between two qubits.
By chaining quantum gates together in the right sequence, you build a quantum circuit — which is essentially a program. The gates are the instructions. They’re important because without them, you have no computation — just idle quantum particles.
And finally, quantum circuits and tensor products:
A quantum circuit is a sequence of quantum gates applied to qubits in a specific order — think of it like a recipe. Each step in the recipe is a gate, and each gate transforms the qubits in a precise way. By the end of the circuit, the qubits hold the result of your computation. You then measure them to read that result out.
Regular computer programs are lists of instructions for bits. Quantum circuits are the same idea, but for qubits — and the “instructions” are quantum gates expressed as matrices.
Now here’s where tensor products become essential. A single qubit is one vector. But real quantum computers use many qubits working together. When you combine qubits into a multi-qubit system, you use the tensor product to build one larger vector that describes the entire system at once.
This matters because quantum circuits often involve gates that act on multiple qubits simultaneously — like entanglement gates. Those gates are matrices built from tensor products too. The whole circuit — every gate, every qubit, every interaction — lives inside this tensor product framework.
So tensor products aren’t just a math trick. They’re the structure that lets quantum circuits scale beyond a single qubit and do something genuinely powerful.