Chapter Ten

Before the Machine Spoke

Kenji Yokosuka did not look like the kind of man most people imagine when they hear the words quantum calculation. That was one of the reasons I liked him.

He was not cloistered and untouchable behind a wall of equations. He was alive in a very ordinary way. His hands usually had some evidence of the physical world on them: a faint line of grease near one knuckle, a half-healed scrape from a stubborn bolt, or the clean little cuts that come from working too long around metal edges. He loved old rock and roll, not as nostalgia, but as proof that ordered noise could still make the blood move. He had opinions about guitar tone, carburetors, timing chains, superposition, entanglement, and whether a car should sound civilized at idle.

His own car did not, even at idle. On the road, it announced him before it arrived.

Not obnoxiously, though some would disagree. Kenji had no patience for loudness without purpose. He liked machines that were tuned, not merely modified. He liked combustion the way he liked mathematics: controlled violence made useful by law.

That was Kenji. A man who could spend Saturday morning adjusting the idle on an old muscle car, Saturday afternoon reading papers on quantum phase estimation, and Saturday night explaining both with the same kind of boyish seriousness.

He was self-taught in the way serious people sometimes are—not as an excuse for ignorance, but because curiosity had refused to wait for permission. He had taken the long way in, through books, lectures, papers, simulations, failed code, online courses, late nights, and more humility than most credentialed men are willing to admit they need.

He respected formal training. He simply had not needed a university to tell him the questions mattered.

That gave him an unusual posture. He did not speak like a man defending his status. He spoke like a man trying not to lose the trail.

When I first began asking whether quantum computation might help us weigh evidence bearing upon design and God as Creator, Kenji did not laugh. He also did not flatter me.

“No,” he said.

That was his first answer. I appreciated it later. At the time, less so. That became one of the reasons I trusted him.

Kenji was never impressed by the usefulness of an answer. If the idea was weak, he would rather disappoint me early than help me defend it badly later.

“No?” I asked.

“No,” he repeated, and turned the wrench in his hand another quarter turn. We were standing in his garage beside the open hood of a car that looked old enough to have stories and loud enough to interrupt them. A small radio near the back wall was playing a guitar riff that had as much confidence as polish. Kenji leaned over the fender, listening to the engine as though it were a witness trying to remember the truth.

“You cannot build a machine that proves God,” he said.

“I know that.”

“I am not sure you do.”

That stung because he was partly right.

He shut the engine off. The garage settled into the ticking heat of metal cooling down.

“You can use computation to compare models,” he said. “You can test consistency. You can ask whether certain patterns are more expected under one framework than another. You can assign priors, define likelihoods, expose assumptions, reduce the weight of weak evidence, and refuse to let enthusiasm outrun method.”

He wiped his hands on a blue shop towel and looked at me.

“But you cannot put the Almighty on trial and ask a circuit to return a verdict.”

That sentence ended the foolish version of the project. It did not end the project.

It saved it.

A Mind Built for Edges

Kenji lived at the edge between disciplines. He had a gift for making difficult ideas understandable without making them smaller. He could take an idea that looked impossible from a distance and find the working parts inside it.

“People make quantum sound mystical because they don’t understand it,” he once told me. “Then other people make it sound boring because they are afraid of sounding mystical. Both are mistakes.”

He was right.

Quantum theory does not need exaggeration. It is strange enough without decoration. Particles do not behave like tiny billiard balls. Measurement matters. Quantum probabilities cannot always be treated as though they merely conceal values we have not yet discovered. Some systems can exist in states that resist the categories of ordinary experience. Entanglement links the measurable states of separated particles in ways that still trouble common sense.

But none of that means quantum theory is magic. Kenji hated that word in scientific conversation.

“Magic means no rules,” he said. “Quantum has rules. They are just not the rules you expect.”

That was how he talked. A little dry, a little amused. He was not trying to make the universe weird. He was trying to understand the kind of weirdness the universe actually permitted.

That mattered because the calculation we were considering had to avoid two opposite errors.

The first error was overclaiming. We could not pretend that a quantum computer could settle metaphysics in one glorious burst of light. That would be theater, not Veritology.

The second error was cowardice. We could not refuse to ask a serious question simply because the answer might offend the reigning assumptions of the age.

Kenji helped me see the narrow path between them.

“Your question is not, ‘Does God exist?’” he said. “That question is too large for the machine and too personal for a machine to answer rightly. Your question is narrower. Better. More honest.”

He picked up a marker and wrote on the whiteboard in the garage:

Under which explanatory framework is the accumulated evidence more expected?

Then he stepped back.

“That,” he said, “a machine can help weigh.”

What the Machine Could and Could Not Do

We spent several evenings defining the boundaries. Boundaries mattered. Without them, the whole effort would collapse into either arrogance or nonsense.

The machine could not generate revelation. It could not know or love truth. It could not repent or pray. It could not stand under the stars and feel small, or look inward and feel accused. It could not recognize the ache of shame or the beauty of forgiveness. It could not tell a grieving father why his son mattered. It could not tell a sinner how to be made clean.

A machine is not a soul. But a machine can process a defined question.

It can test internal consistency. It can evaluate weighted inputs. It can compare models under declared assumptions. It can expose when a conclusion depends too heavily on a fragile premise. It can force the investigator to say what he means before he celebrates what he thinks he found.

That last part mattered most to me. The value of the quantum calculation was not that it could replace judgment. It was that it could discipline judgment.

Kenji called it “keeping our hands off the scales.”

That phrase stayed.

We began with two broad explanatory frameworks.

The first said that reality is ultimately undirected. The universe exists without intention behind it. Laws, constants, matter, energy, life, consciousness, morality, language, and meaning arise through impersonal processes, with no purposeful mind built into reality’s foundation.

The second said that reality is directed. The universe is not an accident at the deepest level. Its laws, intelligibility, fine-tuning, information-bearing life, conscious persons, moral awareness, and hunger for meaning are more expected if Mind stands before matter.

Kenji insisted we state both fairly.

“If you caricature the other side,” he said, “you have not beaten it. You have only beaten your drawing of it.”

That was exactly the kind of warning I needed.

So we wrote the assumptions plainly. No loaded language. No sneering. No devotional shortcuts. The directed framework would not be allowed to smuggle in every Christian conclusion at the start. The undirected framework would not be allowed to dismiss evidence simply because it sounded inconvenient.

Both would stand on the same floor. Then we would ask which one better carried the accumulated weight.

The Evidence Bins

Kenji disliked clutter.

His garage was not spotless, but it was ordered. Tools lived where use had taught them to live. Parts were labeled. Fluids were capped. Nothing sacred about neatness, he said. Neatness just kept stupid mistakes from pretending to be deep problems.

He brought the same discipline to the calculation.

“We need bins,” he said.

“Bins?”

“Evidence categories. Separate them first. Combine them later. If you throw everything into one pile, you won’t know which part is doing the work.”

So we made bins, each representing one line of evidence we had already examined and each asking how strongly that evidence favored one explanatory framework over the other.

Each bin would contain evidence. Each piece of evidence would be described in plain language, assigned a relevance, and given a weight. Each weight would be argued over before the run, not after.

That was important. If we adjusted the weights after seeing the result, we would only be dressing bias in a lab coat.

Kenji would not allow it.

“Pre-register the framework,” he said.

I smiled. “You sound like a clinical trial.”

“I sound like a man who knows people cheat best when they think they are being honest.”

That was one of Kenji’s gifts. He did not accuse human nature in a bitter way. He simply accounted for it.

Priors, Likelihoods, and Humility

The calculation needed a basic structure.

Kenji wanted it understandable enough that a serious reader could follow the logic without pretending to be a physicist. He also wanted it rigorous enough that a critic could not dismiss it as a religious mood board.

“Use Bayes as the skeleton,” he said.

Bayesian reasoning, at its simplest, asks how evidence should update belief. You begin with a prior probability, then ask how likely the evidence is under one explanation compared to another or others. Strong evidence shifts the balance more. Weak evidence shifts it less. Ambiguous evidence should not be made to shout.

That was useful because our investigation had never depended on one lonely fact. It was cumulative. No single piece carried the entire case. What mattered was that independent lines of evidence kept leaning in the same direction.

Kenji was cautious even there.

“Leaning is not proof,” he said.

“I know.”

“Say it again.”

“Leaning is not proof.”

“Good. Now we can work.”

Prior. Evidence. Update.

Then under them:

Declare. Test. Revise.

The simplicity helped.

We would begin with equal priors, not because both frameworks were equally plausible in every philosophical sense, but because equal priors made the test harder to accuse of favoritism. Fifty-fifty. Directed and undirected. Then evidence would be introduced by category, and the model would update according to declared likelihood ratios.

If fine-tuning was much more expected under a directed framework than an undirected one, it would tilt the calculation. If some evidence was disputed or weak, it would tilt little or not at all. If an evidence bin could be explained equally well under both frameworks, it would contribute little. If something favored the undirected model, it would be allowed to move the result that way.

That mattered morally as much as mathematically. Truth has no need to fear honest weights.

That became one of our rules.

Why Quantum?

The obvious question was why quantum computation mattered at all. Couldn’t a classical computer run a Bayesian model?

Yes. For many versions of the problem, it could.

“If we only need arithmetic, use a laptop,” he said. “Do not bring a race car to move a lawn chair.”

That was Kenji again. Cars found their way into his metaphors even when he tried to stop them.

But the calculation we were exploring was not merely a spreadsheet. It involved large model spaces, competing dependencies, uncertainty across domains, and possible combinations of evidential weights. We were not asking one question in a straight line. We were exploring a landscape of possibilities.

Quantum computation, in principle, could help sample or search certain probability landscapes differently than classical methods. It could examine many possible configurations more efficiently under the right structure. It could help test whether the conclusion held across ranges of assumptions or collapsed the moment one weight changed.

That last part was crucial. I did not want a fragile result.

A fragile result would be worse than no result. It would give false confidence to believers and easy ammunition to skeptics. If the conclusion depended on one heroic assumption, then the calculation would not be evidence. It would be decoration.

Kenji’s proposal was more modest and more valuable. We would build the model, define the evidence bins, and assign conservative ranges rather than single triumphant numbers. We would let the system test many combinations within those ranges.

Then we would ask: Does the directed framework remain favored across reasonable uncertainty? Or does the conclusion vanish when the assumptions are loosened?

That was a better question. Not flashy. Not theatrical. Better.

The Garage Whiteboard

The final framework did not come together in a laboratory. It came together in Kenji’s garage. That seemed fitting.

A space heater clicked near the back wall. The car sat under a gray cover, its shape still suggesting impatience. A half-disassembled carburetor rested on a workbench beside a stack of printed papers on quantum algorithms. A guitar leaned against an old amplifier in the corner. The radio had gone quiet by then. Outside, the cold had turned the windows dark enough to reflect us back at ourselves.

Kenji rolled a whiteboard away from the wall and uncapped a marker.

“All right,” he said. “No poetry for ten minutes.”

“That may be hard for me.”

“I said ten.”

He drew two columns: Directed and Undirected. Below them he drew seven boxes: Cosmology, Fine-tuning, Mathematics, Biology, Language, Morality, and History.

Then he drew arrows from each box into a central node labeled cumulative weight.

“No sermons,” he said.

“No sermons.”

“No victory laps.”

“No changing inputs after the result.”

He looked at me.

“No pretending the machine can do more than it can.”

I nodded.

“No pretending the machine can do more than it can.”

He capped the marker.

“Good. Now the machine may be useful.”

I studied the board. It looked almost too simple for the weight it was being asked to carry. But that simplicity was part of its strength. The model did not need to imitate God’s knowledge. It only needed to discipline ours.

Kenji stepped back, arms folded.

“I need one more rule,” he said.

“What?”

“If the calculation favors the directed model, you do not say, ‘Quantum computer proves God.’”

“I won’t.”

“You especially do not say, ‘Quantum computer proves Jesus.’”

“I know.”

“Say what it actually says.”

I waited.

He pointed to the board.

“Say the accumulated evidence, modeled conservatively and tested across uncertainty, favors a reality in which mind precedes matter.”

That sentence was not as exciting as the headline version. It was also far more honest. And in the long run, honesty has more force than excitement.

The Question Beneath the Calculation

The more we worked, the clearer the true issue became.

The calculation was at the heart of the investigation. But it was still only a tool within it. That distinction protected everything.

If the machine gave us a result we liked, it would not become an idol. If it gave us a result we did not like, it would not become an enemy. It would be treated as an instrument. Useful, limited, answerable to truth.

Kenji understood this clearly. For all his love of machines, he did not worship them.

He loved machines because they revealed whether a man had respected reality. An engine does not care about your theory if the timing is wrong. A circuit does not care about your sincerity if the wiring is bad. A quantum model does not care about your desired conclusion if the inputs are incoherent. I understood then why Kenji had been the right man for this part of the investigation.

He loved machines, but he trusted reality more.

That distinction had begun working its way into me too. A tool could strengthen an argument, expose a mistake, or test a method. It could never be permitted to become an authority merely because it was sophisticated.

Machines can be humbling that way. They make poor gods. They make useful servants.

That was how Kenji saw the quantum run. Not as revelation. Not as salvation. Not as a shortcut around repentance, faith, history, Scripture, or Christ.

It was a disciplined weighing—a way to ask whether the accumulated evidence was more expected under intention or accident, and to test Veritology against something more demanding than my own judgment. A way to let the accumulated evidence pass through a structure that would not be impressed by my hopes.

I needed that. I think every investigator does.

Ready to Run

By the end of the night, the garage smelled faintly of oil, coffee, dry-erase marker, and cold concrete.

Kenji had filled the board twice and erased it once. The second version was cleaner. Fewer claims. Better boundaries. Less triumph. More truth.

That was the version we kept.

The run would not happen there. The hardware required more care than a garage could provide, and the final model would need review before it touched any machine worth using. But the essential work had been done.

The question had been narrowed. The claims had been restrained. The weights would be declared. The evidence would be allowed to lean where it leaned.

Kenji stood beside the whiteboard, rubbing at a streak of marker on his thumb.

“You understand what happens next?” he asked.

“We test it.”

“Yes. But more than that.”

I waited.

“We give up control.”

That was not the phrase I expected from him. He smiled a little, seeing that.

“I do not mean spiritually,” he said. “Not exactly. I mean methodologically. Once the model is locked, once the assumptions are declared, once the run begins, we do not get to protect ourselves from the answer.”

The garage was quiet.

The covered car sat behind him like a waiting animal. The whiteboard stood between us, plain and unromantic, carrying the cleanest version of the question we knew how to ask.

Directed or undirected. Mind before matter or matter before mind. A universe expecting life, or life as the fortunate afterthought of indifferent laws.

The question was not small. But it was now honest enough to run.

I closed my notebook.

“All right,” I said. “Let’s build it.”

Kenji nodded once.

Then, because he was Kenji, he reached over, turned the radio back on, and let an old drum solo fill the garage as if the world needed reminding that even noise can become music when the pattern is right.