Bzzzt—
Theo glanced at his cell phone. Then, with an expression as if nothing had happened, he deleted the message.
The sender was “Renaissance Technology.”
It was Wall Street’s legendary quantitative fund, composed of mathematicians. (Quantitative Fund: a fund that makes investment decisions based on mathematical algorithms.)
“What is with everyone lately?”
He grumbled in a low voice so that Seha wouldn’t hear. Recently, headhunters had been contacting him far more often.
Click.
[Invitation to Join the Yale Mathematics Department as a Researcher]
[Regarding Offer for Chief Researcher Position at Google DeepMind]
[Final Selection Notice for Clay Mathematics Institute Fellowship]
[Regarding Recruitment for the Cryptanalysis Division of the U.S. National Security Agency (NSA)]
Offers filled his inbox.
‘Are they settling for the next best thing?’
Theo let out a faint laugh.
What they truly wanted was probably Seha. But he was an extremely difficult target to approach.
He was only sixteen. No matter how much of a world-shaking genius he was, he was not old enough to sign a contract directly. On top of that, the school’s powerful protective measures guarded him.
Minor protection regulations, contract guidelines personally drafted by the head of the legal team, mandatory prior approval for any outsider attempting to contact Seha. Every proposal sent to Seha was double-checked.
So the detour they chose was Theo.
An American team member with high academic credibility, capable of entering into legal contracts, and above all, close to Seha.
“They want to use me as bait.”
With a bitter expression, Theo clicked the delete button.
‘You people don’t know Seha’s true value.’
What they wanted was function.
Genius as a tool to read patterns in the stock market or maximize the efficiency of artificial intelligence. Perhaps even MIT was paying more attention to the economic value Seha would create.
Scratch, scratch.
Theo turned his head toward the sound of chalk.
It was their regular meeting time, and Seha was standing in front of the blackboard. With the hood of his sweatshirt pulled over his head, he was staring fixedly at the equations.
“Theo, this is an incredibly unconventional approach. I thought you could do it, but this exceeds my expectations.”
Seha nodded in admiration.
‘That’s only because you’re here. Even if I don’t have the ability to implement it, as long as I have the concept, I trust that you’ll somehow make it work.’
Theo smiled wryly.
A way to reduce computational load while maintaining accuracy.
His demand was a mathematical contradiction in itself.
Precision and speed were like a seesaw. If one side rose, the other inevitably had to fall—a zero-sum game.
It was the same as telling him to deny head-on the absolute law of signal processing, the Nyquist–Shannon Sampling Theorem. (A theory stating that to convert analog to digital without loss, samples at more than twice the frequency are needed.)
But Theo did not complain. Making the impossible possible was what made a great mathematician.
What would happen if Seha threw himself into this problem? Theo thought that before long, he would solve it as if by magic.
‘No, maybe he already knows the answer.’
Theo applied every theory he could think of to the problem.
Fluid mechanics, cellular automata, chaos theory, game theory, the Poincaré–Hopf theorem, Hamiltonian mechanics. But none of them yielded the answer he wanted.
It meant an entirely new theory was needed, but Theo had no confidence.
‘What if I’m second-rate?’
A passably talented, mediocre mathematician who could never make a name for himself alone, who merely evaluated theories others had created and added footnotes to them.
Crumpled papers filled the room.
They were equations he had written and torn up repeatedly over the past few weeks.
One day, while the frustrating situation continued, as Theo was organizing the photographs he had taken so far, a hypothesis suddenly flashed through his mind.
On the monitor was a photo he had taken on the day he first came to MIT. It was the derivation of Smale’s problem that Seha had written. A high-resolution file of 45 million pixels, an enormous lump of data that, when enlarged, allowed even the dust clinging to the blackboard to be seen clearly.
But even if this were converted to one-twentieth of its size, the meaning conveyed by the photograph would not change.
‘Even if 95% of the pixels are thrown away, information loss is zero.’
In that instant, an idea struck his mind like lightning.
‘Why was I trying to process all data equally?’
Theo hurriedly opened his notebook. Before the idea evaporated, he had to record it. What he wrote on the paper was a short and simple sentence.
‘Information is not equal.’
The sky is empty to a degree infinitely close to 100 percent.
Yet thanks to a few points scattered across it, people were able to find their way even at sea.
“…”
He had formed a hypothesis, but proving it was another matter entirely. No matter how much he thought about it, this was not his field of expertise. Theo decided without hesitation to discuss it with the team members.
And just as he had expected, Seha solved it in an instant.
After hearing the concept, he thought for a long while, then began writing equations on the blackboard.
“Theo’s hypothesis is extremely interesting. If the signal we’re trying to observe has sparsity, then we can drastically reduce the amount of computation. It means that with very few samples, just a few randomly chosen ones, we can perfectly reconstruct the whole.”
Seha picked up the chalk and began writing equations in an empty space on the blackboard.
Scratch, scratch.
In the silence, the crisp sound of chalk rang out.
Theo, Sri, and Sujeong watched the process with bated breath. At Seha’s fingertips, the hypothesis Theo had only vaguely imagined began to take on a mathematical form.
Sri tilted his head as if puzzled.
“Is that a result that can be mathematically guaranteed? To me, it just sounds like a rough conjecture…”
Seha pointed to one side of the blackboard.
“You’re right, it’s a conjecture. A conjecture with an extremely high probability, where the error converges almost to zero.
The way a typical algorithm processes data is similar to how we play Twenty Questions. You ask all twenty questions from beginning to end, then derive the most reasonable answer.”
Sri nodded. That was the traditional Nyquist theorem he had learned, the method of sampling the whole.
“But what if you start knowing that the answer is either a fruit or the name of an occupation?
We can change the questions. And with just two or three questions, we can know the answer. We become able to efficiently eliminate tens of thousands of remaining possibilities.”
Seha gave Theo a thumbs-up.
“Let’s put Theo’s name on this algorithm. It was a brilliant idea. It can be used in every field that requires data.
Theo’s theory will soon become an academic standard.”
His intuition was being realized through Seha’s logic. Theo felt an overwhelming surge of emotion in his chest.
‘I’m not second-rate.’
In Seha’s eyes, he could see trust directed toward him. Theo suddenly thought back to the emails he had deleted earlier.
‘Wall Street? NSA? Clay Institute? Don’t make me laugh.’
This thrill and joy could never be felt anywhere else.
“No, if you hadn’t proven it, it would have ended as nothing more than a delusion.”
With a dazed expression, Theo gave the credit back to Seha. But Seha stubbornly shook his head.
“No. This belongs to you, Theo. How about naming it something like ‘Lanford Compression’? Make sure you publish it as a paper later.”
But Theo also thought it would be dishonest to attach only his own name to it.
It was the moment “ST Compression,” named after the initials of the two of them, came into the world.
* * *
Whirrrrr—
The cooling fans of The Core were screaming. The members of Team Apex watched with tense expressions.
Sri said in a trembling voice,
“Compression algorithm applied. Data density reduced to 5 percent.”
The graph on the screen was fluctuating wildly.
Under normal circumstances, this load would have brought up warning lights and crashed the system, but the data stream to which ST Compression had been applied was passing through the system without resistance.
“Lossless confirmed. Information entropy preservation rate, 99.999%.”
“I can’t believe this is actually possible…”
Sujeong exclaimed in disbelief.
The roaring hum subsided, and the graph regained stability. The runaway flood of data had met the powerful filter called ST Compression and been sifted through.
“Stabilization complete. Resource occupancy is 12 percent.”
Sri muttered blankly as he stared at the monitor.
The engineers had warned Seha. Even if he used the entirety of The Core, it would not be enough to compute all this vast data.
However, the algorithm the team had devised together had dramatically raised the system’s efficiency.
All that remained was to transplant Daisy and proceed with the final test.
* * *
A forty-story building on Thomas Street in New York.
At first glance, it might look like an ordinary building, but this was a place whose true identity was known to only a select few in New York.
Because of its bizarre exterior, with not a single window, this building had long been the subject of all kinds of rumors.
“Ah! There? It was built during the Cold War, so isn’t it obvious? It’s a secret bunker facility for wiretapping communications.”
“I heard it was the NSA’s New York branch.”
“Couldn’t it be something like a weapons research lab? They say compasses stop working if you even go near it.”
Suspicions of this sort could at least be called reasonable. At the very least, they fell within the bounds of reality.
There were also plenty of people who insisted on absurd conspiracy theories, such as it being the headquarters of the Men in Black, a nest of vampires, or a facility for preserving humanity in preparation for the apocalypse.
But the truth was always underwhelming.
This was the communications hub of AT&T, America’s largest telecommunications company.
A fortress-like structure designed during the Cold War to withstand even nuclear war, with thick concrete walls, its own power generation system, and even air purification devices prepared for biochemical terrorism. It was the heart where all of New York’s communication networks intersected.
So in a way, it was only natural that Oracle’s location had been set here.
Basement level 29.
There was an area that not only ordinary people, but even personnel involved, could not enter.
Whirrr—
The elevator came to a stop with a heavy sound.
—Please provide security authorization.
John brought his eye to the camera.
—John Carson, identity confirmed. Clearance Level 2.
When the door opened, cool air flowed out. Along both sides of the corridor, server racks stretching endlessly emitted blue LED light.
It was a computing cluster built with an astronomical budget.
As he stepped inside, a central control room surrounded by massive glass walls came into view. In front of the enormous multi-screen display that filled the wall stood the power brokers who ran New York.
“You’ve arrived, Mr. Mayor.”
The man at the center greeted him.
Seated around him were the commissioner of the New York City Department of Transportation, the police commissioner, the speaker of the city council, and key council members from both parties.
And at the innermost part of the control room, in front of the main console, Seha was seated. He rose from his seat and approached with his team members.
“Hello, Mr. Mayor.”
“Good to see you. You have no idea how long I’ve been waiting for this day.”
Today was the day of Oracle’s demonstration.
For the past few weeks, Seha and Team Apex had been staying here, refining Oracle’s settings. And at last, every preparation had been completed.
“You won’t be disappointed.”
Seha smiled with a confident expression.
John looked around and nodded.
“Now then! Show us. Show us that the tax money of New York’s citizens has not been spent meaninglessly.”
Seha nodded and sat at the main console.
“Oracle, start the system.”
—Understood, Master.
Wooooong—
Along with the sound of the servers running, the enormous screens covering the wall turned on all at once.
Flash!
The moment the screens lit up, exclamations burst out from the room.
The entirety of New York City rose onto the screen, rendered in 3D like an elaborate miniature.
John stared at the sight, at a loss for words.
Equations and graphs, a city map updated in real time, overlapped in layers to form an immense panorama. Traffic flow throughout New York, power consumption, communication traffic, and even pedestrian information based on them.
Within this space, New York was breathing as if it were a single living organism.
“Impressive.”
“Are you saying this can really be processed without any latency?”
“Magnificent. It’s incredible.”
Everyone looked at Seoha as if demanding an explanation.
“This is a digital twin. Another New York, created from data.”
They had only heard of it in words; none of them had expected a system this vast and sophisticated.
A man with the air of a former soldier looked at Seoha and asked,
“I heard Oracle is a prediction-based system.”
“Yes. That’s correct.”
“How accurately can it see the future?”
“It’s not that it can actually see the future, but…”
Seoha let the end of his sentence trail off, then touched the screen.
“It says there is a 67 percent chance an accident will occur at this location in three minutes.”
Everyone was unable to hide their shock at Seoha’s words.
“That’s absurd!”
“You’re saying it can really predict something like that?”
“Enough!”
When John raised his hand, the control room fell silent.
“Seoha, is there a chance it could be a fatal accident?”
“It says there is a high probability it will be a minor accident.”
Seoha answered while looking at the screen.
“Then let’s watch and see. But what is the basis for saying an accident will occur?”
“Oracle? Explain it to them.”
-A bus route change, a signal cycle discrepancy, the driving tendencies of nearby drivers, and one driver’s delayed reaction. At present, there is a high probability that the driver in question is in a physically and mentally compromised state.
The monitor rapidly zoomed in, showing the intersection of Seventh Avenue and 34th Street in Manhattan.
On a road crowded with countless vehicles, a blue sedan was marked in red.
-Target vehicle, unstable driving pattern. Lane departures: three times per minute. Reaction speed: 58 percent slower than the average driver.
Oracle’s dry voice echoed through the control room.
Gulp.
The sound of someone swallowing could be heard.
Dozens of council members and high-ranking officials stared at the screen with bated breath.
-30 seconds prior.
The police commissioner frowned, arms crossed.
“Hmm. Is that car going to suddenly dart out or something?”
-50 meters ahead, City Bus 142 will move into the right lane to enter the stop. The driver of the sedan is currently in the blind spot.
As Oracle explained, the bus on the screen changed lanes.
It was at that very moment. The taxi in front of the sedan sped up, as if refusing to yield.
Vroooom!
But in the end, the moment it was forced back by the bus driver who had already turned the wheel and hit the brakes—
Bang!
The blue sedan on the screen vividly rear-ended the car in front of it.
The cars following behind slammed on their brakes, and the road instantly turned into chaos.
“…”
“My God.”
“Oh, Lord.”
Silence.
An eerie hush flowed through the room.
What broke the silence was the radio at the police commissioner’s waist.
Bzzzt.
Bzzzzzt.
-This is Seventh Avenue Patrol. Collision at the 34th Street intersection. Minor contact accident between a taxi and a passenger vehicle. Fortunately, there are no casualties, but traffic is backing up. Requesting support.