This is Scenario 3. Low revenue pressure — 0.1 out of 1. High technical focus — 0.9 out of 1.

You'd expect a quantum circuit to say build. The technical signal is dominant. The logical answer is go deep on the work.

It didn't say build.

On IonQ Forte-1 — trapped ion qubits — it selected Active Sales Mode at 94% confidence.
On IQM Garnet — superconducting qubits — different hardware, different manufacturer, different country — it selected Active Sales Mode at 92% confidence.

The CNOT gate read the low revenue signal as avoidance behavior. You're hiding in the code instead of making the call.

Two quantum computers. One non-obvious answer. That's not noise. That's a finding.

We ran every scenario twice. Once with quantum entanglement — the CNOT gate connecting the two signals. Once without — each qubit resolving independently.

Under high pressure, the CNOT gate changed the outcome every time. Scenario 1 — strategic planning — flipped from Technical Build to Strategic. Scenario 2 — technical task — flipped from Strategic to Technical Build.

Under low pressure, it didn't matter. Scenarios 3 and 4 came out the same either way.

That pattern held on IonQ. It held on IQM. Two different hardware architectures producing the same split.

Entanglement isn't magic. It's a pressure sensor. When the stakes are low, classical probability is fine. When the stakes are high, the relationship between variables changes the answer. That's the finding.

Scenario 3 has been producing a non-obvious result since the first experiment. Low revenue, high technical focus — and the circuit keeps selecting Sales Mode instead of Build Mode.

The skeptic's question: is that a happy accident? Hardware noise? A quirk of the circuit design?

So we built a test. We flipped which qubit controls the CNOT gate — swapping the dominant variable — and made four mathematical predictions about where the anomaly would move.

All four predictions were correct.

The anomaly moved from Scenario 3 to Scenario 2. Exactly where the math said it would go. Scenario 3 became normal. A new anomaly appeared at Scenario 2.

That's not noise. The circuit is detecting which variable is structurally dominant in the relationship between the two inputs. We named it, we predicted it, and we proved it on real quantum hardware.

Here's the full picture.

Three experiments. We ran them on IonQ Forte-1 — trapped ion qubits operated by laser pulses in a data center in the US. Then we ran them on IQM Garnet — superconducting qubits in Finland, different physical implementation, different manufacturer, different error characteristics.

Every finding replicated.

The Scenario 3 anomaly — the non-obvious result that no classical model would predict — appeared at 94% on IonQ and 92% on IQM. The entanglement proof — high pressure diverges, low pressure converges — held 2-for-2 on both machines. The control flip — all four mathematical predictions confirmed — ran on IQM because IonQ was offline, and still came in 4/4.

When a finding holds across two different physical implementations of quantum computing, you're not looking at hardware noise. You're looking at something in the circuit design that's real.

That's what architecture-independent means. And that's what we built.