The rotors were still spinning down, a descending whine, and the dust was still moving when she lowered her hands.
“Doesn’t look like I have a choice,” she said. “Come inside.” She looked past him at the six units spread across her clearing. “They stay out here.”
He turned and made a brief, flat gesture with two fingers. The units held position.
Inside, he removed the trenchcoat and folded it over the back of a kitchen chair without being invited to, the way a person does when they’ve spent enough time in other people’s homes on Company business that the social calculus has ceased to register. Beneath it he wore a suit that had been expensive once and was still well-maintained — charcoal, no tie, the collar open by one button. He was somewhere in his mid-fifties, tall enough that he’d noticed the doorframe and ducked slightly on his way in, with the kind of build that suggested he’d been athletic at some earlier point and had since replaced that habit with stillness. His face was long and composed, the lines in it reading less as age than as the accumulated sediment of professional neutrality practiced until it became structural. He had fine hands — she noticed this the way she noticed all physical details that didn’t fit the frame: a scientist’s hands, the knuckles unmarked, the fingers tapered. A man who had spent real time at a bench, at some point in his life, before the trenchcoat.
She put the kettle on without asking and stood at the counter with her arms crossed.
“Dr. Veyne,” she said.
“Doctor.” He sat where she hadn’t offered him a seat, at the kitchen table, and set nothing on it — no case, no device, no paper. Whatever he’d come to say, he’d come to say from memory. “To what do I owe the pleasure of the Company?”
“That’s actually my line.”
“Of course.” He looked around the kitchen the way people do when they’re giving you time to settle before they say the difficult thing. His gaze moved across the shelves of specimen jars she kept up here — not the serious ones, the finished ones, the ones she’d brought upstairs because she liked to look at them — and then settled back on her. “We have a problem with N9a.”
She said nothing.
“You know what it is.”
“I know what it is.”
“Then I’ll skip the introduction.” He folded his hands on the table. “The commercial production pathway for N9a — the one the Company has been operating at scale for the past eleven years — runs through a genetically modified strain of phytoplankton. Emiliania-derived, significantly altered, engineered to produce the lipid-encapsulated FePt nanoparticle matrix at concentrations that make industrial harvest viable. The compound binds at the nodes of Ranvier, enables the human-machine interface at the neurological level — you’re familiar with the mechanism.”
She was. She’d been familiar with it for longer than he probably knew.
“The phytoplankton is grown in open-ocean cultivation arrays,” he continued. “Twelve arrays, distributed. Combined surface area of approximately eight hundred square kilometers. It is the only production pathway that meets current demand. There is no secondary pathway. There is no strategic reserve of meaningful size. N9a availability, at current consumption levels, is a function of those arrays remaining productive.”
The kettle reached temperature. She poured without turning around.
“Eleven days ago,” Veyne said, “we identified a Phage variant in Array Seven. By day four it had propagated to Arrays Three and Eleven. Current projections have full array compromise within sixty to ninety days.”
She turned around. “What variant.”
“A recombinant strain. It targets the cellular interface between the modified FePt-synthesis pathway and the phytoplankton’s native lipid metabolism. Specifically, it disrupts the cell wall integrity at that junction — decouples the synthesis pathway from the host cell’s resources without killing the cell immediately, which is what makes it difficult to detect at early stages.” He paused. “The cell wall degradation signature is actually one of the reasons your name came up.”
She set a cup in front of him that he hadn’t asked for and sat down across the table.
“Why don’t you simply let the optimization layer figure it out?”
He picked up the cup. “That would be the obvious question.” He looked at it for a moment. “The layer can optimize known processes. It can model known organisms in known relationships using known datasets. It is very good at that. What it cannot do is find a solution that requires understanding an ecological relationship that was never documented, or draw on traditional practices that were never digitized, or identify an organism that was excluded from the industrial datasets because it had no apparent commercial relevance at the time the datasets were built.” He set the cup down. “The solution to this problem may not exist anywhere that the optimization layer can see. The data it needs may exist only in fieldwork. In observations. In the kind of accumulated ecological knowledge that comes from years of studying systems that everyone else decided weren’t worth studying.”
He looked at her steadily.
“Systems like the ones in your basement,” he said.