The two words were: Carrier Selective.
I read them twice. Set the photograph down.
Turned to Renard.
"Explain," I said.
He picked up his coffee — still barely touched, now cold — looked at it, and set it back down with the expression of a man who had accepted that this particular coffee was not going to be drunk and had made his peace with it.
"Sit down," he said.
"I'll stand."
A pause. He studied me with the careful attention of someone recalibrating how to deliver information based on how the recipient is standing. Then, without further preamble:
"The technology in those patents was not designed to be transmitted randomly," he said. "It was designed to be transmitted to a specific category of person." He paused. "Not a specific individual. A category. A biological profile. What we called, in the project documentation, a Carrier — someone whose neural architecture was compatible with the interface in ways that the general population is not."
I said nothing. I let him continue.
"The interface — what the patents call the miniaturised signal system — was not a device in the conventional sense. It was not something you implanted surgically or wore externally. It was designed to transmit on a carrier frequency that the target neurology would absorb and integrate. Like a signal finding a compatible receiver." He paused. "The Selective part referred to the selectivity of the transmission. Only compatible neural architectures would integrate it. Others would simply not receive the signal at all."
"You designed a technology that would transmit itself into a human being," I said.
"We designed a technology that would interface with a compatible human being," he said. "The distinction matters. An implant is something imposed. This was — by design — a partnership. The interface requires an active neural substrate to function. It calibrates to the specific architecture it encounters. It becomes, over time, something that belongs equally to the system and to the person." He paused. "It is not something we did to you. It is something that happened between you and the technology."
"Because of an explosion," I said. "That you're now telling me was not an accident."
"Because of an explosion that forced a transmission sequence that was not meant to occur in that manner," he said. "In controlled conditions, the process would have been — different. Voluntary. Informed." He paused. "What happened in Corridor B was not how this was supposed to work."
"But it worked," I said.
"Better than any controlled transmission we had achieved in testing," he said, with a quality in his voice that I couldn't immediately name. Something between professional satisfaction and something more complex sitting behind it. "The autonomous recalibration — the system's ability to adapt to its carrier — has proceeded further and faster than our models predicted. Nineteen months. The calibration level you have reached typically required three to four years in our projections."
"Because I've been using it," I said.
"Because you've been using it intelligently," he said. "The system learns from decisions, not from outcomes. Not from whether a trade is profitable — from the quality of the analytical process that precedes the decision. You provided it with exceptional data." Another pause. "You are, James, the most successful carrier integration we have ever observed. And we have been observing for fourteen months."
Étienne, who had been standing by the kitchen counter with his arms crossed and the expression of a person trying to process information that his existing categories were inadequate to contain, said: "You've been watching him."
Renard looked at Étienne briefly. "We've been monitoring the market outputs, yes. Not James directly." Back to me. "Your trading record was how we confirmed the integration had succeeded. The Vinci position. TotalEnergies. The Sasung short — which, I should tell you, caused considerable alarm on our end when we saw it. Not because of the trade itself, but because of the probability score it implied."
"Eighty-nine percent," I said.
He went still.
"You know the exact figure," he said.
"I know all the figures," I said. "Nineteen months of them."
He looked at me with something that was genuinely not what I had expected from a man who had arrived at Étienne's apartment at seven in the morning to recover a proprietary technology. He looked at me with something that was almost — not quite, but almost — respect.
"The eighty-nine percent suggested the system had achieved a calibration level we did not think was possible in under two years," he said. "It suggested access to data synthesis inputs that our theoretical models placed at year three or four of carrier integration." He paused. "Which was when we realised you were not an ordinary carrier."
"You said I wasn't an accident," I said. "Explain that part."
He set his hands flat on the table.
"The facility roster on the day of the explosion," he said. "The people present in the building. Every person was accounted for in the incident report — staff, researchers, contractors. All of them were screened in the weeks after, as part of the standard post-incident protocol. Screened for integration." He paused. "Every person tested negative. The interface had not transmitted to any of them."
"Except me," I said.
"You were not on the roster," he said. "You were not a staff member, not a contractor, not a registered visitor in the system. You were carrying coffee for a researcher who had not logged your presence." A pause. "When we reviewed the CCTV footage from the corridor — before the explosion destroyed the cameras — we found you. An unregistered visitor. A seventeen-year-old boy carrying a cardboard tray." He paused. "We had never screened for you because we did not know you existed."
"And when you found me?"
"By then, your integration was three months advanced," he said. "Screening at that stage would have been — disruptive." He paused. "And then your trading record began to emerge, and we understood that disruption was no longer the relevant concern."
I looked at him.
"You left me alone," I said. "For fourteen months. Knowing I had it. Knowing I didn't know what it was."
"We monitored," he said. "We assessed. We determined that intervention before you understood what you had would be counterproductive." A pause. "And honestly, James — watching what you did with it was the most complete field test our technology has ever had."
Something moved in my chest. I identified it after a moment as anger — quiet, specific anger, the kind that didn't require expression because expressing it would accomplish nothing useful.
"I was your field test," I said.
"You were our proof of concept," he said. And then, with something in his voice that I did not expect: "And you exceeded every parameter we had set."
I looked at him for a long moment.
"Who are you?" I said. "Not Renard. Who do you work for?"
"The team that built the system," he said. "We were a division within Sasung's European research arm. When the facility was destroyed, we — dispersed. Some of us to other companies. Some of us to independent work." He paused. "I have been trying to locate and protect the surviving outputs of our work. You are the most significant surviving output."
"I'm a person," I said.
"Yes," he said, without hesitation. "You are. Which is why I'm here talking to you rather than reporting your location to the people who destroyed the facility."
The apartment was very still.
Étienne had gone completely motionless by the kitchen counter.
"Who destroyed it?" I said.
Renard was quiet for a moment — the pause of someone deciding how much of a truth to give.
"People who believed that the technology should not exist," he said. "Not for moral reasons. For competitive ones." He paused. "A system that generates real-time probabilistic outputs at the calibration level you have reached is not simply a research curiosity. It is a market instrument of extraordinary power. The people who destroyed the facility did not want the technology to exist because they could not control it." He paused. "They are not regulators, James. They are not governments. They are private actors with very large positions in the markets the technology can predict."
I thought about Édouard Aldric. Seventy-one years old. Twenty-two years without a loss. Forty million in Sasung Europe.
"Someone who had a lot to lose," I said.
"Several someones," he said. "A consortium. Informal but coordinated. They identified our research eighteen months before the explosion and made a decision." He paused. "They did not succeed in destroying the technology. They succeeded in destroying the facility and killing two of my colleagues." His voice remained even on those last five words, the evenness of someone who had processed grief for fourteen months and arrived at something on the other side of it that was not peace but was functional. "And they do not yet know the technology survived."
"You said," I said carefully, "that someone made a trade that should have been impossible. In the chapter teaser—"
I stopped.
That was not a thing I had said aloud. That was a thought.
Renard looked at me with something that was now entirely undisguised.
"The system is sensitive at high calibration levels," he said. "Particularly to the carrier's concentrated attention." A pause. "I did not hear a thought, James. But I noticed the moment you began constructing one. There is a — quality — to high-calibration carrier cognition that the monitoring equipment we have operational picks up as a signal variance." He paused. "You were thinking about the chapter teaser."
I set that aside entirely.
"The trade," I said. "What trade gave it away?"
"Your Sanofi position," he said. "Entered April second. The probability output that informed it — 79%, we calculated, based on your position size and risk parameters — incorporated data from an EMA advisory committee pre-meeting document that was distributed to committee members forty-eight hours before your entry." He paused. "Pre-meeting documents are not public. They are circulated only to committee members under embargo."
"The panel accessed an embargoed EMA document," I said.
"The system's data synthesis inputs include regulatory databases that carry a forty-eight-hour advance access protocol," he said. "It was a design feature. For the system to generate meaningful probability scores, it required data with sufficient lead time to be actionable. We built access protocols into the infrastructure." He paused. "We did not anticipate that those protocols would survive the facility's destruction."
"They're still running," I said.
"The infrastructure was cloud-distributed," he said. "Destroying the physical facility did not eliminate the data access layer. The system — through you — is still drawing on inputs that were designed for controlled research use." He paused. "Which means the AMF investigator who reviewed your Sanofi trade was not wrong to be troubled."
I thought about Chassagne. About his face in Brasserie Georges. About "impossible" — the word he had used off the record to Petra Novák.
He had been right. Just not in the way he thought.
"So the people who destroyed the facility," I said. "How do they know the technology survived?"
"They don't," Renard said. "Not yet. But the Sanofi position — the EMA pre-meeting document — it generated a flag in the data access log. A non-standard access pattern from a carrier-side query." He paused. "The access logs are monitored. Not by us. By the infrastructure provider." He paused again. "And the infrastructure provider's parent company has a board member who is connected to the consortium."
"So the flag was seen."
"It was logged," he said. "Whether it was understood — that depends on whether someone in that organisation knows what they're looking at." He paused. "I estimate we have between two and six weeks before that flag is interpreted correctly."
Two to six weeks.
Before the people who had destroyed the facility, killed two researchers, and spent fourteen months believing they had succeeded — before they understood that the technology was operational and in a person and that person had been trading European equities from Lyon for nineteen months.
"What do you want from me?" I said.
"What I want," Renard said carefully, "and what I'm asking for are different things." He paused. "What I want is to understand the full calibration state of the system — to know what it has become after nineteen months of your particular kind of use. What I'm asking for is something simpler." He looked at me. "I'm asking you to let me help you. Because when the consortium finds you — and they will find you — you will need to know what you are dealing with."
"And what am I dealing with?" I said.
"People who destroyed a facility to prevent a technology from existing," he said. "Who will view an integrated carrier operating at your calibration level as a significantly greater threat than a research project." He paused. "You have turned their mistake into a weapon, James. And they are going to want it back."
The morning had moved on while we had been talking — the light in the apartment was different now, the city fully awake outside the window, the ordinary Tuesday proceeding without any awareness of what was being said on the second floor of a building on Rue Pasteur.
"I need to speak to my lawyer," I said.
"Of course," Renard said. He reached into his jacket again and produced a card — plain, no company name, just a mobile number. "When you're ready." He stood. Straightened his jacket with the unhurried precision of a man who had delivered what he came to deliver. "The system chose you for a reason, James. Not randomly. Not incidentally." He paused at the table, looking at the photograph of Corridor B. "The Carrier Selective protocol screened for specific cognitive profiles. Pattern recognition at the high end of the distribution. Risk calibration — the ability to act on incomplete information without paralysis. And a particular quality that our researchers debated for years about how to name." He paused. "We called it orientation. The condition of having a direction so clear and so fixed that it provides structural coherence to every decision." He looked at me. "The system found you because of the person you already were. Not the other way around."
He picked up the photograph and placed it in his jacket pocket.
He left.
The door closed behind him.
Étienne and I stood in the apartment as the sound of his footsteps descended the stairs and the building door closed two floors below with its characteristic sound.
Then Étienne said, in the voice of a person for whom language had once again become temporarily insufficient:
"James."
"I know," I said.
"There are people who blew up a building—"
"I know."
"—and killed people—"
"I know."
"—and they're going to find out that the thing they blew up the building for is—" He stopped. "James, it's in your eye."
"I know," I said.
He looked at me. At the particular face I knew I was making — the controlled, sequential face, the face I made when I was processing something that had not yet been fully processed but would be, given time and patience and the right kind of thinking.
"Are you okay?" he said.
Nobody had asked me that in nineteen months. Not in the way he asked it — not as a social formality, but as a genuine question about my interior state from a person who cared about the answer.
I considered the question honestly, which was something I did not always do.
"Not entirely," I said. "But I will be."
He nodded slowly.
"What do we do?" he said.
"You do nothing," I said. "Today. Nothing that involves the fund, Beaumont, Berenger, or any of this. You go to your lecture. You have lunch. You live your day."
"And you?"
"I call Maître Dubois. I think." I picked up my jacket from the back of his chair. "And then I do what I always do."
"Which is?"
I paused at the door.
"Make a plan," I said.
I called Maître Dubois from the street. Told her about Renard — not everything, but the shape of it. The carrier selective protocol. The data access layer. The consortium. The two-to-six-week window.
She was quiet for fifteen seconds after I finished, which was longer than she had ever been quiet.
"Come to the office," she said. "Now."
"I need an hour," I said. "There's something I need to do first."
Another pause. "James—"
"One hour," I said.
She accepted it.
The library was quiet at ten in the morning. The serious students had claimed their tables by eight; the casual ones wouldn't arrive until noon. The space between those two populations was the most productive hour of the day in this building, and I had known it for years.
I found a table near the east window and opened my notebook.
Not to write a plan. Not yet. The plan required information I didn't have and resources I needed to assess and a conversation with Maître Dubois that would shape everything.
I opened it to a blank page and I sat with it.
I was doing this — the sitting, the deliberate pause before action — because Renard had said something that I was still turning over. Orientation. The condition of having a direction so clear and so fixed that it provides structural coherence to every decision.
The system had chosen me because I had direction.
My grandfather's forty-three euros. The plan that had been running since I was twelve years old, accreting detail and structure and patience in the nineteen months since the corridor, building toward something that I had always been able to see even when it was very far away.
That direction was still there. The consortium had not changed it. Renard had not changed it. The revelation of what the panel actually was had not changed it.
But it had added weight to everything.
I was still looking at the blank page when someone set a cup of coffee on the table beside my notebook.
I looked up.
Margaux stood beside the table with her own cup, a book under her arm, and the expression of someone who had done something and was waiting to see whether it had been the right thing.
"You look," she said, with characteristic precision, "like someone who is trying to think through something that is too large to think through in one sitting."
I looked at the coffee. Then at her.
"How did you know I was here?" I said.
"I didn't," she said. "I come here on Tuesday mornings." She paused. "May I sit?"
Under any other circumstances, I would have said yes immediately — Margaux was useful and perceptive and had, in a series of brief encounters across several weeks, demonstrated a quality of intelligence that I found, if I was being accurate, genuinely rare.
But this was not any other circumstances.
This was a Tuesday morning on which I had been told that the thing in my eye was a stolen technology, that people had died to prevent it from existing, and that the people responsible had somewhere between two and six weeks before they found me.
"Another time," I said. "I'm sorry."
She looked at me for a moment. At whatever my face was doing — the controlled face, the sequential face, the face that Étienne had learned to read and nobody else had been given enough exposure to decode.
She didn't press. Didn't ask why. Didn't make the moment uncomfortable by reacting to the refusal with anything that required management.
She simply nodded, once, and moved to a different table.
And that — the ease of it, the complete absence of social demand in her response — was the thing that made me look at her twice.
Not because it was convenient. Because it was rare.
I had met very few people in my life who understood instinctively that sometimes the right thing to give someone was space. Who didn't interpret a boundary as a rejection that needed to be processed aloud. Who simply accepted the shape of a moment and moved within it without friction.
Étienne was one. In a different register, in a different way.
Margaux was another.
I looked at her for a moment longer than I had intended.
She was reading. Her book open, her pen moving in those small precise strokes, her attention entirely redirected as though the previous thirty seconds had simply not occurred.
I looked away.
I picked up the coffee she had left on my table. It was the right temperature — not scalding, not lukewarm. The right amount of milk. Which meant she had watched me make coffee in this library before without my noticing.
I filed that carefully.
Then I closed my notebook, picked up my jacket, and went to see Maître Dubois.
At the door of the library I stopped and looked back once.
Margaux was still reading, her pen moving, her attention complete.
I left.
But the coffee was still warm in my hand when I reached the street, and I did not set it down until I had finished it, which was something I noticed only after I had done it.
Maître Dubois listened to everything without interruption for twenty-two minutes.
When I finished, she was quiet for eleven seconds.
Then she said: "The Beaumont claim is no longer our primary problem."
"No," I said.
"The consortium," she said. "If they find you before we understand who they are—"
"They won't," I said. "We have the window."
"Two to six weeks."
"Yes."
"James, two to six weeks to identify and neutralise a private consortium that has the resources to destroy a research facility and has been operational for at least two years is not—"
"We don't need to neutralise them," I said. "We need to make finding me less valuable than leaving me alone."
She paused. "Explain."
"They destroyed the facility because the technology threatened their market positions. Because a system that generates real-time probability scores at this calibration level disrupts whatever edge they currently hold." I paused. "But destroying me doesn't destroy the technology anymore. The calibration is in me — in my neural architecture. It can't be extracted and it can't be transferred except through a process Renard's team designed and the consortium doesn't have." I paused. "So their options are eliminate me or accept that the technology exists."
"And if they choose elimination—"
"Then I need to be visible enough that my elimination is costly," I said. "Not to me personally. To them institutionally." I paused. "The best protection I have is the same protection I've always had."
She looked at me.
"The fund," I said. "JE Capital. Berenger. The AMF file. Chassagne. Every institutional connection I have built in nineteen months creates a web of attention around me that makes me difficult to disappear quietly." I paused. "An unknown student in Lyon can be disappeared. A registered fund manager with an institutional LP and an open AMF file and a compliance lawyer cannot."
She studied me for a long moment.
"You want to grow faster," she said. "Not for the money. For the visibility."
"I want to become someone whose absence would be noticed," I said. "By enough people, in enough institutions, that making me absent becomes a more expensive problem than tolerating my existence."
A silence.
"That," she said, slowly, "is either the most sophisticated protective strategy I've heard from a client or the most dangerous." She paused. "It requires JE Capital to grow significantly in a very short window. Which means more trades. More results. More attention from exactly the people who are already watching."
"Yes," I said.
"Chassagne will notice," she said.
"Chassagne is already watching," I said. "And he gave me his personal number and his word that he'd warn me before he acted." I paused. "He's watching for a different reason than the consortium. The AMF wants to understand me. The consortium wants to eliminate me. Those are not the same kind of attention."
She was quiet.
"Margaux Villeneuve," I said.
She looked up. Slightly surprised by the name — the first time I had ever mentioned her.
"The Berenger research intern," she said. "I've seen her name in the filing records."
"She identified that Étienne would crack under sustained professional pressure," I said. "She told me I needed a better human structure — someone who understands enough to cover what he can't." I paused. "She was right."
"You're considering bringing her in."
"I'm considering," I said. "Not deciding. Not yet." I paused. "She's perceptive and she's discrete and she understands markets in a way that Étienne doesn't. She's also not — she doesn't—" I stopped.
Maître Dubois waited.
"She doesn't push," I said. "For information she hasn't been given. For access she hasn't been offered." I paused. "I find that unusual."
Maître Dubois looked at me with the expression of a person seeing something that had not previously been visible.
"You trust her," she said.
"I don't know her well enough to trust her," I said. "I'm noting that the conditions for trust are present in a way they rarely are."
She was quiet for a moment. Then, with the precise professional warmth she occasionally allowed herself: "That, James, is the most personal thing you have ever said to me."
I said nothing.
"Go home," she said. "Sleep. You've been awake for over thirty hours." She picked up her pen. "Tomorrow we build the plan. Tonight you rest."
I stood.
"The consortium," I said. "Two to six weeks."
"I know," she said. "I'll make calls this afternoon." She paused. "James. The system chose you because of your orientation. Your direction." She looked up at me. "Don't lose it in all of this. The direction is the asset. Everything else is circumstance."
I nodded.
I left.
On the walk home, the city going about its afternoon with complete indifference to everything that had happened in it since seven that morning, I thought about the coffee.
The right temperature. The right amount of milk.
Someone paying that kind of attention, quietly, without announcement, without expectation of acknowledgement.
I thought about the blank page in my notebook.
About orientation. Direction. The fixed point that structured everything.
I had always known what I was building toward.
I was beginning to wonder, for the first time, whether what I was building toward was only what I had named it.
End of Chapter 11
The Probability of Wealth — By J. Valmont
[Next Chapter → Chapter 12: "Seventeen Days"]The consortium's window closes faster than Renard predicted. James has seventeen days. He makes five trades, grows the fund to eight million, and calls Margaux Villeneuve for the first time — not about the fund.