The mint-green Mini Cooper was waiting at the corner of Madison Avenue. As soon as I clicked the seat belt into place Tabitha popped the clutch and shot into the traffic.
“The whole world seems to know about Chuck’s new toy pointing the finger at Harry Nash,” I said.
“FinCEN techs wired it up. He should have known. But that’s not what’s really bothering him. Chuck has friends in the West Wing,” she said. “Apparently they have private polling that says the president’s headed for a train wreck in the next election.”
That would account for Chuck’s hesitation to open a file on Nash.
We went through the park and took the West Side Highway downtown.
“The computer guy, is he there now?” I asked.
“He’s waiting.”
“And he’s worked for us before?”
“The Russian diamond-smuggling network. He wrote that program for tracking the diamond flows from the Russian mines so we could tell how much Lime was laundering for the oligarchs.”
I remembered him now. Patrick Ho. PhD from Stanford in computer science. He had worked for the Office of Tailored Access Operations, the team of NSA hackers that penetrated the Chinese military’s Shanghai-based hacking operation. For weeks, the team at Fort Meade had logged every keystroke that the Chinese made in their campaign to penetrate the US government’s computer defenses. Finally, when they’d got all the information they wanted, Tailored Access had snapped pictures of the Chinese hackers, using the cameras on the hackers’ own computers. The first the Chinese knew they’d been hacked was when Tailored Access posted their faces online, with their names.
Patrick had later joined a unit at the Treasury that was modelling some kind of big-data weapon to help the United States bludgeon other countries with the power of the dollar. I mean, more than we were bludgeoning them already. Tabitha had managed to get him temporarily seconded to me when we launched the original operation against Lime.
Her window was open. Her hair flew around her head. The car smelled of her shampoo.
“Patrick Ho,” I said. “His dad is Ho Wang Wei. Used to run a big mahjong game in Chinatown.”
“He still does,” Tabitha said.
That was true, but since I’d never put it in a file Tabitha must have developed sources of her own. Strictly speaking, she should have told me first. I filed it away in the corner of my mind where I store suspicion. It’s a large corner.
We turned off the highway into a parking lot beside one of the cavernous old freight terminals that survive on the west bank of Lower Manhattan. They stick out into the Hudson like fingers clawing at the past. Around them the waterfront blazes with glass condos and outdoor restaurants and parks where people walk their tiny dogs. Against this, the old piers mount a crumbling rearguard of boxing gyms and doomed businesses that ebb and flow like the murky waters of the river.
As we got out of the car, I waved to the black man who was sliding the chain-link gate shut behind us. Augie Treacher ran the pier. He did me favors because I was Tommy’s friend. Tommy had arranged a no-contest plea deal with suspended sentence after Augie had walked into a Queens restaurant with a rivet gun and fatally interrupted two members of the Capezi crime family in the middle of the pasta special, which that night was spaghetti carbonara. Two months before, they had cut out Augie’s tongue in a dispute about who said what to whom.
“And you’ve explained the problem?” I asked Tabitha as we climbed a narrow flight of stairs.
“He understands,” Tabitha said.
“And his clearance?”
“It’s current.”
The narrow passage smelled of damp metal and oil and old timbers. The steel stairs echoed as we climbed. On the top floor, cracked linoleum added the odor of old wax to the smell of the river. At the end of the hall was a tiny office cut in half by a wooden counter. Behind the counter hung a rack of golf clubs. I grabbed a driver, and we pushed through a steel door onto the driving range.
After the dim interior the light was blinding. Across the river the towers of Hoboken blazed in the sun. The driving range stretched before us, 350 yards of Astroturf surrounded by nets hung from pylons. Beyond the netting, white hydrofoil ferries went skimming down the river to the Battery. In front of us a tanker the length of a football field plowed slowly up the Hudson.
We stepped onto the driving platform. I placed a ball on a tee and handed Tabitha a driver and stood back to watch the most beautiful swing I had ever seen that was not on the PGA. The ball lifted in a lazy arc, following a perfect line. It bounced onto the green carpeting around the 225-yard marker. We’d been here before. We both knew that I would follow with a vicious slice, so I got it over with while Tab tried not to look.
Patrick Ho was waiting in the glassed-in booth to one side of the tees, where the golf pro kept his bookings, his whisky, and his despair. I don’t know why he was always so morose. If he wanted to remind himself that things could be worse New Jersey was right there across the river.
Augie had closed the range for a few hours, and the rickety pro shop was a handy place to sit. The glass rattled and wind whistled through cracks and water lapped at the old black piers. I shut the door behind me and sat down.
Patrick Ho was short and compact, with a crewcut and black-framed glasses. He wore a gray turtleneck, black chinos, and brown suede ankle boots, an attire that managed to be monastic and stylish at the same time, as if he belonged to an order of monks that ordered its habits from J. Crew.
“Tabitha says you think you can help us.”
“Maybe. The problem, as she explained it, goes roughly like this. Nash bought into First Partners when the Russian oligarchs forced Lime to sell. Later, Nash bought the Russian Pink from the fund. Question: Are Lime and Nash in some kind of partnership, and is the Russian Pink part of it?”
“That’s right. I don’t know why Nash would buy shares in a Russian fund and then immediately buy an asset from it. But he had a reason, and I think it’s a reason he shared with Lime.”
“Do we know how First Partners bought the pink? I’m assuming it’s not a normal kind of business acquisition for a fund.”
I told him that all I had was a tip from a South African source that Barry Stern, a Johannesburg dealer, had paid $12 million for a large pink.
“If we take that as the start of the money trail, can you follow it?”
“Maybe later, when we have more data,” Patrick said. “But there’s something else I’d like to try first.”
The wind was rattling the glass so hard I was leaning across the desk to hear him. We left the pro shack and made our way down to the end of the range. A Circle Line boat loaded with tourists wallowed through the waves.
“Shoot,” I said.
He hooked his fingers through the netting.
“It’s something I’ve been thinking about for a long time. I’d like to create a program that thinks like the person we’re targeting.”
“And how do you do that?”
“By using a new kind of computing system called a neural network. The problem with most computers is that they only know how to do what we tell them to do. They don’t think, they follow rules.”
He warmed to his plan. As he explained it, a computer following rules never benefits from its experience. It doesn’t get smarter with each successive operation. A neural network actually got smarter. Through a process called machine learning, the computer absorbed data and learned to draw conclusions from it.
“But how do you know its answers are right?”
He took off his glasses and polished them carefully. He held them up and squinted at them before he put them back on.
“Do you ever know someone is lying to you? You have no proof, but you’re certain anyway?”
“Sure.”
“That’s because many people have lied to you, and you’ve discovered they have, and now you’re able to recognize the lie as it’s happening. You don’t follow rules, you’ve just learned to know. A neural network works like that.”
He used medicine as an example. Neural networks were already making complicated diagnoses. Instead of writing rules to tell the computer what a lung tumor looked like, programmers created data sets of hundreds of thousands of images and scans of lung tumors. They fed them into the machine. The computer learned what lung cancer looked like the way a doctor learned: by seeing it.
“The computer looking for a tumor knows what it’s looking for,” I said. “This time it won’t.”
“It will learn,” Patrick said. “It will sift through emails and phone logs and bank records, and eventually, if I’m right, it will figure out how Lime did business, not only with Stern but with Nash.”
A bright red chopper from the heliport near the Battery appeared in the sky and headed for the Statue of Liberty with another load of tourists. Two hundred bucks apiece for a seven-minute spin. Huddled masses, welcome to New York.
“Earlier you mentioned computers we have access to. You’ve probably gathered that what I’m asking you to do is off the books. That’s why we’re meeting here. You wouldn’t have a Treasury contract.”
“I know that,” Patrick said.
“I want to make this clear. When I said off the books, I meant way off.”
Patrick nodded, took off his glasses, and polished them again with the bottom of his shirt.
“You cut my dad a break. There were some aspects of his business that were not one hundred percent legal.” He held his glasses up to the sky again. “If you had gone strictly by the letter of the law, my dad would have been deported as a felon. And anyway,” he said with a sudden grin, “the Treasury will never catch me. When I go into the network, they won’t know I’m there.”
Even with the river lapping at the pier and the deep thrum of a tugboat as it chugged upriver, the ping from my phone sounded loud.
I could count on one hand the people who have my cell number. Unrecognized sender—right away that’s bad. No message, just a link to a live stream. It took me a moment to recognize what I was looking at, and when I did a block of ice formed around my heart. It was a stretch of highway I knew well. That’s how the horror starts. Something mundane. It’s your life, and now someone else is driving it.
That highway was the road to Annie’s school.