“Not so easy to be an ethical robot,” Flanagan conceded. “I should be taking notes.”
“Yes, you should.” She smiled. “Have you ever heard of ‘The Trolley Problem’?”
“Can’t say that I have. Why does the trolley have a problem?”
“My goodness. You really are an engineer!”
“Yes, ma’am. Bugs and plugs.”
“The Trolley Problem assumes that a runaway trolley is about to hit five people who are tied to the tracks. If you throw a lever, the trolley will switch to a siding where only one man is tied down. Are you visualizing this? Okay. What’s the ethical robot supposed to do?”
“Stop the trolley.”
“That’s not an option, unfortunately. The question is whether you make the choice and kill one, or let things happen, and kill five.”
“One. Duh.”
She nodded and tapped her glass with one of those lacquered nails.
“That sounds right. And ninety percent of people give that answer. But to see the problem with that decision, you need to think about ‘Fat Man on a Bridge.’ You probably don’t know that one either. You interested?”
Flanagan was trying to keep his distance, but her intensity and clarity made it hard to stand apart.
“Yes, Denise, I’m interested.”
“Okay. Our train is heading down the tracks toward those same five people, but this time you’re on a bridge above and you can stop the train only if you throw something heavy in its path. There’s a fat man standing next to you. Should you push him off the bridge so he stops the train and saves the five people, even though he’ll end up dead?”
“Does it matter that he’s fat? Suppose he’s a thin man?”
“He has to be fat. Otherwise he won’t be heavy enough to stop the train. It could be a fat woman, I suppose, but that introduces other issues. So yes, a fat man.”
“Push him,” said Flanagan.
“You’re the exception,” she said with a wink. “It turns out that most people who were willing to kill the person who was tied to the tracks to save the other five don’t want to push the fat man off the bridge. Isn’t that interesting?”
“What would you do?” asked Flanagan.
“Push him,” she said. “Maybe it’s something about our line of work.”
Flanagan held up his white ceramic cup for the flight attendant to signal that he wanted a third helping of mixed nuts, but the attendant said it was time for lunch.
“Let’s go back to robots,” said Flanagan. Despite himself, he had gotten interested in the conversation. “Could someone hack an ethical robot and make it a bad robot?”
“Of course,” she said quietly. “I believe some of our colleagues are working on that very problem now. The adversary’s robots are presumed to be unethical. That’s one of the unquestioned assumptions around here, if you hadn’t noticed.”
Flanagan leaned a bit closer. He lowered his voice almost to a whisper.
“Could you recruit an ethical robot to spy for you?”
“Well, well.” She looked away for a moment, out the window to the tabled farmland below. “That’s an interesting question. Recruit the robot! I’m not sure. If the robot was ethical, could it allow itself to be reprogrammed? I need to ponder that one.”
Flanagan waited for more, but she was rustling in her seat, adjusting the pillow.
“I think I’ll have the pasta,” she told the attendant. But she barely touched it.
She dozed off after lunch. Flanagan went to the bathroom and recorded on “Voice Notes” as much as he could remember of what she had said. But he wondered, as he whispered the words into the phone, whether he was just helping her cement a studiously crafted cover story.
They rented a car at LAX and traveled down the 405 to Newport Beach and Orange County’s version of the Riviera. Flanagan drove and Ford fell asleep in the passenger seat. It was a crystalline fall afternoon; Flanagan savored the ride and hoped his companion wouldn’t wake up.
They checked into a fancy hotel in Newport Beach near the headquarters of a big bond-trading firm. Ford said she wanted a comfortable venue for the world-class computer scientists they would meet the next day. She was quiet, talked out; she said she was going to order a salad from room service and get a good night’s sleep.
“I hope you read some of those books I gave you on quantum computing,” she said before parting. “You’re going to need them tomorrow.”
Flanagan promised it would be his bedtime reading. He set a monitor across the hall from her door to alert him if she left. She didn’t stir until morning.
Flanagan took a run along the beach late that afternoon, as the sun was disappearing into the Pacific. He drove to a stylish restaurant in town and sat at the bar with people who looked like movie stars, and maybe they were. He had a martini—a “see-through,” as his Boston Irish father always called them—and a pork chop. He had struggled during the flight to LAX to find the telltale admission in Ford’s smart, sinuous talk and failed. Now, his mind was pleasantly empty.
27.
NEWPORT BEACH, CALIFORNIA
Denise Ford convened her meeting with contractors the next morning in a conference room overlooking the sea. The sign outside the door said simply “Computer Science Group.” A hotel manager wheeled in a cart with coffee, hot water, soft drinks, and pastries and then promised to leave them undisturbed.
Ford was gracious and demure. She wore a black suit with a white blouse and low heels. Her eyeglasses were perched on the bridge of her nose.
“Welcome, all,” she said, rapping a coffee cup with her spoon. The morning sun was bleaching the little room. She held the center of the table easily, effortlessly, looking at each guest and gaining assent as she framed the agenda.
“Thanks for joining us this morning. Before we start, some housekeeping: This conversation will be unclassified but under the non-disclosure agreements that you have made previously. If each of you could please sign the forms at your place, acknowledging that you are aware of the rules.”
Ford paused while the grant recipients signed their forms. When they were done, Flanagan gathered them up and she put them in her briefcase.
“So on to business. It may seem strange to sign agreements that you can’t disclose your requests for greater openness, but welcome to the world of government contracting. I know from our correspondence that you all share concerns about classification and publication rules for your quantum computing projects, going forward. Right?”
Voices assented around the table.
“I am here to listen to your concerns and answer them, if I can. I have brought along my new deputy, Mark Flanagan, who, like me, is a longtime agency employee. He will help me report back your comments to our colleagues.”
Flanagan introduced himself to the group. He apologized that he was a mechanical engineer, rather than an electrical one, but he promised to take careful notes.
Ford introduced the three scientists sitting at the table. Howard Sagan, a professor at the University of California at San Diego, was developing a new approach to building stable, long-lived “qubits.” He looked like an advertisement for the California good life, tan and fit, wearing a black t-shirt under his blazer. Next to him was Carson Malloy, the chief scientist of a big computer company that was financing Sagan’s research. The third attendee was Andrea Bildt, a blonde woman in her thirties who ran a start-up that was developing new programming languages.
Malloy was the spokesman for this insurgency. He had been designing and building supercomputers for thirty years. He was the only person in the room, other than Flanagan, wearing a suit and tie. He was precise and emphatic.
“We’re doing world-changing work on quantum computing thanks to Howard and Andrea,” he began. “Breakthrough work! We’re making so much progress, in fact, our biggest worry is that you’ll insist that it be classified. That would be a terrible mistake. Research is an organic process. It will thrive in the sunlight and wither in the dark
. That’s why we asked for this meeting. We respectfully must protest!”
Malloy leaned back in his chair and looked over the top of his glasses. Sagan and Bildt nodded vigorously.
“Hold on, folks,” said Ford, raising her hand gently. “Before we go to the barricades, you need to explain what this ‘breakthrough’ is about. Then we can talk about rules. And I remind you: I don’t make these decisions. I may not even agree with them. But I’m here to listen, on behalf of the government. So tell us: What do we need to know?”
Howard Sagan, the professor, answered for the group. He had a PowerPoint presentation queued up on his computer, and he now projected it on a screen at the foot of the table. The first slide showed a familiar picture of Albert Einstein, with his electrified shock of white hair, and next to him, a round-faced man in aviator glasses and a bushy moustache.
“Let’s begin by recalling what’s at stake when we talk about quantum computing. Every schoolchild knows Professor Einstein, the founder of modern physics. But for us in this room, the more immediate ancestor is the other gentleman on the screen, Peter Shor, professor of physics at MIT. He taught us in 1994 that if we could build a quantum computer, it could factor numbers much faster than a classical computer. How much faster? Let’s remind ourselves.”
Sagan’s next slide showed a computer chip above two mathematical formulas for factoring large numbers. The first portrayed a classical computer, whose required calculations would be proportional to ten raised to the power of the number of digits factored, divided by two. The second showed the formula for a quantum machine, whose steps would be proportional to the number of digits squared—a vastly smaller number.
“What would this mean in practice?” continued Sagan. “We currently estimate that it would take a classical computer more than ten million years to factor a fifty-digit number, whereas it would take less than a second on a quantum computer. No existing cryptographic system could withstand such computer power. All codes could be decrypted and read.”
“We know that,” said Ford. “That’s why we’re here.”
A new slide filled the screen. “QC May Change Many Aspects of Human Life,” read the caption. Sagan summarized the applications that would be transformed by this wondrously fast computer: materials science, chemistry, pharmaceuticals, process optimization, pattern recognition, machine learning. To hear him, the future of global health and happiness would be altered by this magical box.
The slide dissolved into one with a quote from Arthur C. Clarke: “Any sufficiently advanced technology is indistinguishable from magic.”
“Understood,” said Ford. “It’s a digital cornucopia. But to repeat: What’s your breakthrough? That’s the question.”
Sagan flipped ahead several slides. He stopped at one marked “De-coherence.”
“Coherence is the problem for quantum computing, as we have all discovered. We have learned that the wondrous, potentially life-transforming qubits in our quantum computers are very fussy. They’re vulnerable to heat, magnetism, any kind of energy. They often survive only a few milli-seconds. Even the supposed breakthrough with the ‘ion-trap’ approach will not provide an answer to the problem of de-coherence, we believe.”
“We’re not going to talk about the ion-trap research.” Ford wagged her finger. “Not in this session.”
“We don’t think it will work,” broke in Malloy. “And believe me, we know a lot more about this than your ion-trap friends at IARPA.”
“No comment,” said Ford. “Let’s stick to your research for now, not other people’s.”
Sagan projected a new slide, marked “The Secret of Topological Qubits—Braiding.” It showed something that looked like a Mobius strip. Sagan stood now, looming over the image on screen, as he tried to explain his breakthrough.
“We are convinced that the best way to create a truly stable, error-resistant qubit is by braiding quantum quasi-particles around each other so that they are entangled in a quantum state. These quasi particles hide quantum information just as a cipher hides classical information, and this encryption protects the quantum information more precisely than other architectures. Do you perhaps, possibly understand what this means?”
“No,” said Ford. “But you seem to, which is enough for now. But you realize what this means: If it’s as good as you say, then my colleagues in the United States government are going to want to take it off the grid so that you can work on it. Quietly. Safely. Secretly.”
“But that’s a dreadful idea!” exploded Malloy. “You can’t keep this secret. It’s already public. We’re a global company. Everyone knows what we’ve been doing.”
“But they don’t know that it works!”
“This is crazy,” muttered Sagan. “It’s a flat world. You can’t fence things off anymore. We are working with labs all over the world. Some of our closest collaborators are in the Netherlands and Denmark. I have members of my lab from Qinghua University in Beijing and Moscow State University. They’ve moved their families to San Diego. I can’t just tell them all to go home.”
“Yes, you can. If we tell you to. Those are the rules.”
“We don’t want to play anymore, then,” said Malloy. He was glowering at the two CIA officers. “It’s knowledge. It’s free. That’s America’s comparative advantage, for god’s sake. The fact that we’re open. That’s why we get the best grad students from Russia and China.”
Andrea Bildt had been silent until now, watching her male colleagues spar with Ford. But now she spoke up. There was a hint of a Swedish accent in her voice. That was where she had started in her travel across the flat world to Silicon Valley. Her blonde hair was knotted in a braid. She looked as if she might have just come in from surfing at Redondo Beach, a few dozen miles up the coast.
“Let me explain my work,” said Bildt. “Maybe then you’ll understand why you can’t keep this technology in a box.”
“At last, a woman’s voice,” said Ford. “I am tired of arguing with alpha males.”
“Here is what I do, Miss Ford: I am writing a quantum programming language. Strange stuff. But to someone like me, it is poetry.”
“Brava,” said Ford.
“I began with a language called F Sharp. It’s a very high-level language, good for pattern recognition and meta-programming. But then we began to think, suppose Carson and Howard could build their quantum computer. How would we tell it what to do? We would have to write the instructions in braids to match the qubits. So that is what I have written—a braided language that can provide instructions to our entangled qubits. Does that make any sense?”
“Yes and no,” said Ford. “I can’t follow the science, but what makes sense is that you’re telling me you did it.”
“Yes, I did do it. I’m sharing it now with programmers around the world. They want to help. For people like me, this is a very big challenge because nothing in quantum software can be copied, and all operations have to be reversible. Do you see what a challenge that is? It’s hot. Everyone wants to work on it. If we try to stop them now, they will be upset. And if the world hates us, they won’t want to work with us. That would be very bad.”
Ford looked at Flanagan, who had been silently tapping at the keys on his laptop, taking notes. Her look said: See? I told you so.
“You make a powerful case. Especially you, Andrea. I’ll report what you said. But it’s not my call. These rules have been around since the Manhattan Project. They’re not likely to change now. I’m not saying I agree with them. I’m just telling you what it is.”
“Let me show you one more slide,” said Sagan. “Maybe it will help you understand the mystery here. And why you shouldn’t try to choke it with secrecy.”
On the screen appeared a photograph of a handsome young man in a double-breasted jacket and tie. His dark hair was neatly combed. He had a thin, delicate mouth and piercing dark eyes.
“This is Ettore Majorana,” said Sagan. “He was an Italian theoretical physicist, a Sicilian, who studied w
ith Enrico Fermi, then with Nils Bohr, and also with Werner Heisenberg. For a physicist like me, Majorana is a god: Fermi compared him to Galileo and Newton.”
The screen dissolved, and the picture of the young Italian was replaced with a page of dense equations.
“Why is this man important to us?” continued Sagan. “First, he gave his name to the subatomic particles that we are trying to braid in our topological qubit. They are called ‘Majorana Fermions,’ because he predicted them in 1937. These particles are their own anti-particles. They behave the way that entangled qubits do, except that they are more stable. I know that sounds very abstract, but technologists have actually manufactured these strange fermions, at the end points of tiny nanowires that are cooled almost to absolute zero. They are impossible, but real! That is the first reason we revere Majorana.”
“What’s the second reason?”
The screen dissolved once more, and it now showed a picture of a slightly older Majorana, his eyes haunted, in a picture accompanying an Italian newspaper story with the headline “Chi l’ha visto.” Who saw him?
Sagan’s voice lost its quick, academic cadence. He spoke slowly, somberly.
“Majorana disappeared at sea in 1938. He was traveling by boat from Palermo to Naples. He vanished without a trace. His body was never found. The day he left, he sent a note to the director of the Naples Physics Institute. I can quote it from memory. ‘Dear Carrelli, I have made a decision that has become unavoidable. I realize what trouble my sudden disappearance will cause you and the students. I beg your forgiveness.’ And then he was gone. What happened to him?”
“Maybe he fell off the boat.”
Sagan shook his head. Malloy put his finger to his lips.
“Many people think it was suicide,” said Sagan quietly. “Or that he fled, perhaps to Argentina. Or that he escaped to a monastery. But why did he take these radical measures? What frightened him? Many of his friends thought they knew the answer. When he went to work with Heisenberg, the Nazis had come to power. People were already thinking of war . . . and weapons. Majorana did not want his brilliant work in physics to be used by Hitler and Mussolini. He did not want science to be militarized. So he vanished.”
The Quantum Spy Page 23
