Renard found her in eleven hours.
Not because he was exceptional at finding people — though he was — but because Elise Hartmann, it turned out, had not been hiding. She had been waiting. In an apartment on Rue de Rive in Geneva's Eaux-Vives district, fourth floor, with a view of the lake and a laptop open to a document she had been adding to, intermittently, for fourteen months.
Renard called me at ten forty-seven the following morning with the address.
"She knows you're looking for her," he said. "I made discreet contact through a mutual colleague from the faculty. She responded within the hour." A pause. "James. She asked one question when I told her I had found the carrier."
"What did she ask?"
"She asked if you were well," he said. "Not whether the calibration was stable. Not whether the system was functioning. Whether you were well." He paused. "I thought you should know the distinction."
I filed that carefully.
"I want to call her directly," I said. "Not through you."
A beat of silence. "I expected that," he said. He gave me a number.
I called from the back office of Maître Dubois's building, door closed, at eleven-fifteen.
She answered before the second ring.
"James Smith," she said. Not a question. The tone of someone who had been expecting a specific call for a specific amount of time and had simply been waiting for the phone to demonstrate it.
Her voice was lower than I had expected — measured, with a precision that reminded me of Maître Dubois but was different in register. Where Dubois was structured, Hartmann was dense. The difference between a legal brief and a research paper. Both precise. Different kinds of precision.
"Dr. Hartmann," I said.
"Elise," she said. "Please."
"Elise," I said. "I want to ask you directly — the Hartmann injunction. The thirty-two pages of technical analysis. Did you write them voluntarily?"
A pause.
Not the pause of someone constructing an answer. The pause of someone deciding whether to be entirely honest with a person they had never met but had, in a specific and unusual sense, been connected to for fourteen months.
"No," she said. "And yes."
"Explain," I said.
"No," she said, "in the sense that I was given no meaningful choice. They have leverage — professional, legal, personal. The kind of leverage that a consortium with four hundred billion euros and twenty years of practice accumulates with precision." She paused. "Yes, in the sense that when I wrote the analysis, I made choices about what to include and what to omit. The thirty-two pages describe your system accurately enough to establish legal standing. They do not describe it completely." Another pause. "If I had wanted to give them everything they needed to replicate it, I could have written three hundred pages. I wrote thirty-two."
I absorbed this.
"You sandbagged it," I said.
"I gave them enough to file a claim," she said. "Not enough to build one."
"Why?" I said.
She was quiet for a moment.
"Because the system was not built to be owned by people like them," she said. "It was built for something specific. Something Renard has probably told you in his version, which is accurate but incomplete." She paused. "May I ask you something before I tell you the rest?"
"Ask," I said.
"The panel," she said. "In nineteen months of use — has it ever shown you a probability for something that was not a financial instrument? Not a market position. Anything else?"
I thought about this.
"No," I said.
"Has it ever initiated on its own?" she said. "Outside of your directing attention to a specific investment?"
I thought about the hospital. Day three. The panel filling without prompting.
▣ THIS IS ONLY THE BEGINNING.
"Once," I said. "In the hospital. Four days after the explosion."
"What did it show?" she said.
I told her.
A silence.
"That was not the system," she said. "That was the initialization sequence. A self-check — the equivalent of a device confirming successful transmission and basic function." She paused. "The message was pre-programmed. It was not a probability output."
"It was a message," I said.
"Yes," she said. "From us. From the team. To whoever the system found." She paused. "We did not know who that would be. We built the message for a person we had not yet met." Another pause. "We wanted them to know, from the beginning, that what had happened to them was the beginning of something. Not an accident. Not a malfunction. A beginning."
I sat with this for a moment.
"The carrier selective protocol," I said. "Renard told me it screened for specific cognitive profiles. Pattern recognition. Risk calibration. Orientation."
"Yes," she said.
"He said I was not an accident," I said. "That the system found me because of who I already was."
"That is true," she said.
"But the explosion was not planned," I said. "We were not in controlled conditions. The transmission was forced by a destruction event, not a designed delivery."
"Also true," she said.
"So the system found me because I happened to be the nearest compatible person when it was forced to transmit," I said. "Which means I was both chosen and accidental simultaneously."
A silence.
"Yes," she said. "That is the most accurate description I have heard." A pause. "James. That is exactly what Renard's version does not fully explain. Because Renard believes the system chose you. He is a scientist — he looks at the outcome and infers design. But the design was the protocol. The specific individual was — fate, if you believe in it. Probability, if you don't." She paused. "You happened to be in that corridor. You happened to be compatible. And you happened to be the kind of person who, when something inexplicable appeared in his eye, wrote five rules in a notebook and went to work."
"How do you know about the five rules?" I said.
A pause.
"The system's recalibration logs," she said quietly. "They transmit to the infrastructure layer. Renard has access. He shared the early logs with me when I — when contact was first established." A pause. "I have been reading your logs for eight months, James. Not the trades. The decision quality. The analytical process. The moments when you chose to act and the moments when you chose to wait." She paused. "I know you better than you might be comfortable with."
I thought about what that meant. That the infrastructure layer — the cloud-distributed system that was still running, still processing, still providing data access — had been logging not just the outputs but the process. The quality of the thinking that preceded each decision.
"The consortium's acquisition objective," I said. "If they get the infrastructure access, they get the logs too."
"Yes," she said. "Which is not the part that worries me most."
"What worries you most?"
"The recalibration is ongoing," she said. "The system continues to develop. In nineteen months you have taken it further than our models projected for year three. If the consortium acquires the infrastructure and the technical specification—" She paused. "They cannot take it out of you. The calibration is neural — it is yours and only yours, it cannot be transferred or extracted without destroying it." She paused again. "But they could attempt to create a new carrier. Someone compatible. Someone they control." She paused. "With the full technical specification and the infrastructure access, it would take them three to five years. But they would have a version of what you have. Under their direction."
A three-to-five-year timeline to build a controllable version of the panel.
Under the direction of a consortium managing four hundred billion euros.
I thought about Chassagne's word. Impossible. About Petra Novák's analysis. Nineteen percent annualised alpha. About the Efficient Market Hypothesis and the foundational fiction that kept capital markets functioning as though they were fair.
A consortium with a controllable version of the panel would not use it to make markets fair. They would use it to make their own positions infallible while everyone else operated in the dark.
"Elise," I said. "Can you leave Geneva?"
A pause.
"The leverage they hold is not physical," she said. "I am not imprisoned. I am — obligated. They have documents. Professional. Legal." She paused. "One personal matter that I will not discuss." A pause. "If I leave Geneva without resolving the obligation, they will use what they have."
"Tell me what they have," I said.
"Not yet," she said. "I need to know you first. Not the calibration data. You." She paused. "Renard knows the system. He does not know the carrier. Those are not the same thing, and for what comes next, the carrier matters more."
"What comes next?" I said.
She was quiet for a moment.
"They will contact you within seventy-two hours," she said. "Not through Beaumont. Not through Hartmann. Directly. Someone from the Foundation." A pause. "They will make an offer. It will be structured to appear reasonable — a licensing arrangement, a consultation fee, a technical partnership. The language will be designed to make refusal seem irrational." She paused. "James. When they make the offer, do not refuse it immediately."
"Why not?"
"Because the offer is the only moment in which you have leverage," she said. "While they believe negotiation is possible, they will not move to the alternative. The moment you refuse, the alternative begins." She paused. "Use the offer to learn everything you can about who is making it and what they actually want."
"And then?"
"And then," she said, "we will have had this conversation, and you will know what I know, and we will decide together what to do with it." She paused. "If you're willing."
I looked at the closed door. At the sliver of light underneath it from the main office where Dubois and Chassagne and Margaux were working.
"Elise," I said. "The night of the explosion. Renard told me the relay was modified. That it was not an accident." I paused. "He told me the consortium destroyed the facility."
"Yes," she said.
"Was he there?" I said. "When it happened?"
A silence.
A different kind of silence from all the others. Longer. Carrying more.
"James," she said carefully. "What exactly did Renard tell you about the night of the explosion?"
"That the consortium sent someone to destroy the facility," I said. "That two researchers died. That the technology was supposed to be destroyed with it." I paused. "That is what he told me."
Another silence.
"And the relay," she said. "The modified relay in Corridor B. Did he tell you who modified it?"
My hand tightened on the phone.
"He said the DREAL report categorised it as potentially consistent with deliberate alteration," I said slowly. "He implied the consortium was responsible."
"The DREAL report is accurate," she said. "The relay was deliberately modified." She paused. "James. The consortium did not send anyone to the facility that night. Their method of suppression was economic and legal — acquisition attempts, regulatory pressure, intellectual property challenges. The physical destruction—" She stopped.
"Elise."
"The physical destruction of the facility," she said, very carefully, very quietly, "was not carried out by the consortium."
I said nothing.
"The consortium arrived afterward," she said. "They found an opportunity in the aftermath — acquired the remnants, pressured the survivors, began the systematic suppression. But they did not create the opportunity." She paused. "Someone else did."
"Who?" I said.
A silence.
"That," she said, "is the question I have been sitting with in this apartment for fourteen months." She paused. "And the reason I wrote thirty-two pages instead of three hundred. Because until I know who destroyed the facility and why — until I understand whether the destruction was meant to prevent the system from existing or to force an uncontrolled transmission — I cannot be certain of anything." She paused again. "Including whose side Renard is on."
The back office was very still.
"You don't trust him," I said.
"I trusted him for six years," she said. "Before the explosion. After—" She stopped. "After, I found that I could not locate him for three months. He appeared in my life eight months ago, through careful and indirect contact, with a story about tracking carriers and monitoring the consortium." She paused. "It is a coherent story. It may be true. But coherent and true are not the same thing, and I have had fourteen months in a Geneva apartment to think about the difference."
I thought about Renard. About his card. About the eleven seconds of silence before he told me Elise was in Geneva. About the careful way he had described the consortium — as though he knew them from the outside.
About the access logs. The recalibration data. The infrastructure layer that he said he monitored.
An infrastructure that was still running.
That someone had to be maintaining.
"Elise," I said. "The infrastructure. The cloud-distributed data access layer. The recalibration logs." I paused. "Who maintains it?"
A silence.
"It was designed to be self-maintaining," she said. "Autonomous. The physical facility could be destroyed and the infrastructure would continue." She paused. "But for the data access protocols — the regulatory databases, the advance access — those require periodic authentication. Someone with the original credentials." She paused. "Someone who was on the research team."
"How many people had the credentials?" I said.
"Three," she said.
"Rousseau and Herrera are dead," I said.
"Yes," she said.
"You," I said.
"My credentials were revoked when I was pressured into the Hartmann arrangement," she said. "Standard security protocol — I revoked them myself, before they could be used against me." She paused. "Which means the infrastructure is currently being maintained by one remaining set of credentials."
I already knew the answer.
But I needed her to say it.
"Renard," I said.
"Renard," she said.
The silence that followed was the longest of the conversation.
"He has been maintaining the system that lives in your eye," she said quietly. "For fourteen months. Providing it with data. Allowing it to calibrate. Watching it develop." She paused. "James. I do not know if he is protecting you or using you. I do not know if those are different things from his perspective." She paused again. "But I know that whoever maintains the infrastructure controls what the panel sees. Which means they influence, at least partially, what probabilities you receive."
I sat with this.
The panel. Nineteen months of probabilities. Twenty-six positions. An eighty-one percent win rate that had made Chassagne use the word impossible and Petra Novák restructure her emerging manager thesis.
All of it filtered through an infrastructure maintained by a man I had met in Étienne's apartment and given tentative trust to.
"The probabilities could be curated," I said.
"Yes," she said.
"He could have been directing my trades," I said.
"Not directing," she said carefully. "The carrier selective protocol gives the carrier autonomy — the system cannot override your decisions. It can only provide inputs. But inputs can be curated. Specific data emphasised. Other data deprioritised." She paused. "You would always have been making genuine decisions. But the information those decisions were based on—"
"Could have been shaped," I said.
"Yes," she said.
I thought about the Sasung short. Eighty-nine percent. The highest signal I had ever received. The trade that had made Chassagne say impossible. The trade that had put JE Capital on the map.
The trade that had triggered the consortium's attention.
Which had brought Renard to Étienne's apartment.
Which had brought me to this phone call.
"Elise," I said. "Is it possible that the Sasung short was designed? That someone wanted me to make that trade, at that time, at that size, to produce exactly the visibility that followed?"
A very long silence.
"Yes," she said. "It is possible."
"And if it was designed," I said. "What was the purpose?"
She was quiet for a moment.
"There are two possibilities," she said. "The first: someone wanted you visible enough that the consortium would find you, leading to the current situation, for reasons I cannot yet explain." She paused. "The second: someone wanted you visible enough that the consortium would find you — and then that someone planned to use you as the instrument through which the consortium's entire structure is dismantled."
I looked at the door again.
At the sliver of light.
"Which possibility does Renard represent?" I said.
"That," she said, "is what I need to know. And what I need your help to find out." She paused. "James. Are you willing to trust me? Not Renard. Not the system. Me."
I thought about the thirty-two pages written to give the consortium legal standing without giving them the means to build. The question about whether I was well. The fourteen months in a Geneva apartment, waiting.
"Yes," I said. "On one condition."
"Name it."
"Come to Lyon," I said. "Not because I need you here — because you shouldn't be there alone anymore." I paused. "You said the leverage isn't physical. Which means you can leave."
A very long silence.
"What they hold," she said slowly. "The personal matter I said I wouldn't discuss."
"Yes."
Another pause.
"My daughter," she said. "She is seven years old. She lives with her father in Lyon. He does not know about the consortium. He does not know what I built or what was in the facility." She paused. "They found her eight months ago. Not a threat — not explicit. But they sent me a photograph. Her school. The playground." She paused. "That was all. No words. Just a photograph."
The back office was very still around me.
I thought about a researcher sitting in a Geneva apartment for fourteen months. Writing thirty-two pages instead of three hundred. Revoking her own credentials. Waiting.
Protecting.
"Elise," I said. "Her name."
A pause.
"Sophie," she said.
"We will get Sophie," I said. "Before you come. That is the first thing we do. Not the consortium. Not Renard. Sophie first." I paused. "Can you trust me on that?"
A silence.
Different from all the others.
Not analytical. Not careful.
Just — human.
"Yes," she said.
"Give me twenty-four hours," I said.
I opened the door.
Four people looked up — Dubois, Chassagne, Margaux, and Gérard who had arrived an hour ago with the quarterly filing amendments and had not left because the room had reached a critical mass that made leaving feel wrong.
I stood in the doorway for a moment.
"There is a seven-year-old girl in Lyon," I said. "Her name is Sophie Hartmann. I need to know her school, her address, and the fastest way to put someone trustworthy between her and a consortium that has sent her mother a photograph of the playground." I paused. "After that, I have something to tell you about Renard."
The room was very still.
Then Chassagne closed his folder and said: "I know a family protection unit. They owe the AMF a significant professional favour."
Maître Dubois picked up her phone.
Margaux looked at me from across the table with the expression I still didn't have a complete name for — but was getting closer to.
"James," she said.
"Yes."
"Are you alright?" she said.
It was the second time in twenty-four hours that someone had asked me that question and meant it.
"Not entirely," I said. "But I will be."
I sat down.
We went to work.
End of Chapter 15
The Probability of Wealth — By J. Valmont
[Next Chapter → Chapter 16: "Sophie"]Sophie Hartmann is collected from school by a family protection unit at three-fifteen on a Tuesday afternoon. She does not know why. Her father does not know why. And at four-oh-seven, while James is explaining to the room everything Elise told him about Renard and the infrastructure and the curation of the panel's inputs — Renard walks through the front door of Maître Dubois's building. He was not invited. He knew where to come anyway.