One Qubit

Below Vela, the far instrument rotated like a piece of dark glass set into the orbit of Sable. The moon turned irregularly beneath it, a captured stone that had never settled into the courteous motion expected of native satellites. No habitation ring softened the station's outline. Its spin section was buried in the shadow of its own shield. Its laboratories extended on trusses far enough apart that one refrigeration failure would not warm the others.

Five people lived there. They grew mushrooms in the warm return lines, voted monthly on which delayed broadcasts to download, and measured a bad year by how many replacement parts had to be machined from something else. Every personal message shared bandwidth with target telemetry. Every cup of water crossed the same accounting boundary as the vacuum pumps.

The station existed to prepare small physical systems, transform them under controlled fields, and measure what survived the encounter. Its current calibration would let Catenary's traffic network separate beacon phase drift from the charged-particle weather approaching the inner orbits. That result, rather than the elegance of the machine, was why Vela had crossed the reach.

Iona met the crew at the inner lock. She wore no uniform, only a gray work coat with conductive thread bright at the cuffs, and looked first at the evidence case in Sana's hand.

"The Archive signed it?"

"The Archive recorded it," Sana said.

"Better. Signatures make them sentimental."

A slight change in Iona's expression suggested that the answer restored an older confidence. During the instrument's first winter, a corrupted parts manifest had left her with three thousand fasteners of the wrong alloy and a beautifully signed explanation. She had corresponded with Sana ever since, mostly by attaching increasingly specific questions to rejected records.

Only then did she look at Osei. He opened a padded case and offered two bearing sensors, one built for the mounting he remembered and one for the revision installed after he left. Carrying both had cost Vela mass Mara could have sold to cargo.

Iona selected the revised part. "You checked."

"Eventually."

Her thumb remained against the sealed case for a moment before she passed it to a technician. No embrace would have told Tala as much about the years between them.

Iona led them down a corridor whose local down shifted as the station turned. Frost feathered one wall behind a warning grid. At the laboratory entrance they exchanged ship clothing for clean layers and passed through two fields that removed the dust Sable lifted even in vacuum. One shielded device occupied the center of an otherwise empty room.

Its controlled system had two distinguished measurement outcomes. The instrument labeled them 0 and 1. Those labels formed the computational basis, written |0> and |1> when they named quantum states rather than classical integers.

In Dirac notation, the vertical mark and angle bracket formed a ket. |0> was read "ket zero." It did not claim that a tiny written zero lived inside the device.

Wheeler represented the system with one quantum register.

quantum class KnownZero {
  state long measured = 0;
  qreg q = new qreg(1);

  entry void main() {
    prepare(q, 0);
    measured = measure(q);
    assert(measured == 0);
  }
}

qreg q = new qreg(1) declared a register containing one qubit. prepare(q, 0) established the known basis state |0>. The measurement produced one classical outcome and stored it in measured.

KnownZero (quantum) halted
measured = 0
measurements = [0]

A single run established one outcome. The device made no celebratory sound. A green boundary appeared around the record, and the refrigeration plant continued its low structural hum.

Target interfaces called one complete preparation, operation, and measurement trial a shot. Iona requested 256 shots, each of which created fresh state and a complete execution. Measuring the already measured system 256 times would not have been the same experiment.

wheeler run tutorial/KnownZero.wbc \
  --target ideal \
  --shots 256 \
  --seed 271828

0 | ################################################################ 256
1 |                                                                  0

Under the ideal model, this preparation and basis predicted certainty. Hardware could later depart from the ideal through noise and imperfect control, but no uncertainty had yet entered the semantic experiment.

Mara recognized the next operation before Tala did. X exchanged the two basis states just as the finite flip had exchanged classical 0 and 1.

quantum class BasisFlip {
  state long measured = 0;
  qreg q = new qreg(1);

  entry void main() {
    prepare(q, 0);
    X(q[0]);
    measured = measure(q);
    assert(measured == 1);
  }
}

On the computational basis, the table was familiar.

That resemblance had a boundary. A classical bit always occupied one of its allowed values in the classical model. A qubit required more states than the two basis labels could name.

Iona introduced H, the Hadamard operation, without defining those states in advance.

quantum class HadamardSample {
  state long measured = 0;
  qreg q = new qreg(1);

  entry void main() {
    prepare(q, 0);
    H(q[0]);
    measured = measure(q);
  }
}

Before the next run, Iona asked for predictions. Mara expected the two bars to balance exactly because the target was ideal. Osei expected both outcomes but would not name counts. Sana wrote only the scope of the record they were about to create. Tala, still remembering the long count, predicted variation around equal frequencies.

She ran 1,024 fresh shots with a declared seed.

wheeler run tutorial/HadamardSample.wbc \
  --target ideal \
  --shots 1024 \
  --seed 314159

0 | ################################################### 509
1 | #################################################### 515

Mara conceded the six-count imbalance without complaint. Every shot still ended in one classical outcome. The histogram approached equal frequencies, yet nothing in a single outcome identified the state that existed before measurement. Calling the qubit a hidden coin would reproduce the counts and fail the next experiment.

During a cooling pause, Iona took them to the station galley. Five place settings had become nine by borrowing dishes from the machine shop. Osei opened one cabinet without asking and found spare filters where cups had once been. Iona pointed to the opposite wall. His chipped blue cup remained in the common rack, worn by use rather than preserved. She filled it and set it at the place beside hers.

Sable rolled across the ceiling window, its surface black except where old prospecting charges had exposed pale veins. The residents asked about Catenary's gardens and whether the Archive still served the impossible broth. Nobody asked whether the experiment had worked. At the instrument, that question was considered too broad for a meal.

Only the ideal simulator could expose its mathematical state before measurement. That diagnostic belonged to the simulator, not to an ordinary hardware observation.

basis   amplitude
|0>     +0.7071067811865475
|1>     +0.7071067811865475

Each row carried an amplitude. The state after H on |0> was therefore written

(|0> + |1>) / sqrt(2)

or, more compactly,

H|0> = (|0> + |1>) / sqrt(2)

A state with nonzero amplitudes on more than one basis state was a superposition in that basis. The term named the amplitude description. It did not mean that measurement would report two outcomes in one shot.

Probability came from squared amplitude magnitude. Both real amplitudes had magnitude 1 / sqrt(2), so each basis outcome had probability 1/2.

|1 / sqrt(2)|^2 = 1/2

Because the state was normalized, its probabilities summed to one.

1/2 + 1/2 = 1

With the amplitude table in place, the path arithmetic from the incoming packets found its quantum use. Iona applied H a second time before measurement.

quantum class HadamardReturn {
  state long measured = 0;
  qreg q = new qreg(1);

  entry void main() {
    prepare(q, 0);
    H(q[0]);
    H(q[0]);
    measured = measure(q);
    assert(measured == 0);
  }
}

For output |0>, two contributions reinforced.

+1/2 + +1/2 = 1

For output |1>, opposite contributions canceled.

+1/2 + -1/2 = 0

Measurement therefore returned 0 with certainty in the ideal model. A hidden fair coin flipped twice would not have made the same prediction. The amplitudes, including their signs, carried information absent from ordinary probabilities.

Osei studied the zero amplitude. "The route disappeared."

"The route was never an outcome," Iona said. "Its contribution canceled at that destination."

He rubbed the bridge of his nose, an admission he offered more readily than agreement. Then he amended the diagnostic note himself. Such a distinction mattered whenever a diagram tempted someone to assign a classical private history to each quantum path.

A second gate made phase visible. Z preserved |0> and negated the amplitude of |1>.

Applied directly to prepared |0>, Z changed no prediction. Applied between two Hadamards, it changed everything.

quantum class RevealPhase {
  state long measured = 0;
  qreg q = new qreg(1);

  entry void main() {
    prepare(q, 0);
    H(q[0]);
    Z(q[0]);
    H(q[0]);
    measured = measure(q);
    assert(measured == 1);
  }
}

After the first H, the amplitudes were (+,+). Z changed them to (+,-). The final H converted that relative sign into a certain basis outcome, 1.

Negating both amplitudes instead would have produced (-,-), a global phase change. Every amplitude would rotate together, so no interference experiment on the isolated state could reveal the difference. Negating only one row changed relative phase, which later operations could convert into different outcome probabilities.

Beyond two signs lay every other possible phase. The far instrument existed because such angles could carry structure worth measuring. Its oldest clock array still used mechanical shutters cut during the settlement years. Its newest target controlled fields by instructions whose timing would have been meaningless to those builders. Both systems depended on comparing where one cycle stood relative to another.

Wheeler's Phase operation could rotate an amplitude through any declared angle. At a quarter turn, the real number line became a plane.

1       points right
-1      points left
i       points up
-i      points down

A complex number supplied two real coordinates on that plane. Its magnitude measured distance from the origin, while its angle carried phase. Rotating an amplitude could change its angle without changing the probability obtained from squared magnitude.

Tala drew one final state table.

Nothing in the table made complex numbers mysterious. One arrow pointed right, another up, both had equal length, and a later operation could bring their angular difference into interference.

When the calibration ended, station night had come without darkness. The lab lights dimmed while Sable's horizon continued its uneven passage beyond the shield cameras. Iona opened the adjacent enclosure. A second qubit waited inside, coupled to the first by a controlled operation.

On Sana's evidence case, the old Bell record changed status from inscrutable source to a question they could finally approach. The program had been included in the mission package as a target diagnostic. Its lone result, 3, had seemed almost childishly small beside the distance required to produce it. Now Tala could see how much state that integer concealed.

Iona had already opened the next chapter, Two Systems.