Keeper of the algorithm.., p.12
Keeper of the Algorithm (The Keeper Saga Book 1), page 12
Li stood stone-faced.
Zhang muttered quietly, "Their response could be . . . significant."
He cast his eyes down.
Li turned to the screen displaying the intrusion vectors, concealed through layers of proxy servers. "We've been meticulous in our defense. But even if they do respond, we will have achieved the 'probabilistic advantage'."
Zhang looked properly chastened.
Major Wang said, "I've already installed my agents. They've been briefed. I have absolute faith they will execute their mission flawlessly, sir."
Li nodded.
Lieutenant Chu said, “Our surveillance systems are in place to monitor the response, sir."
Li said, "Prepare the package."
Several hours later, they reconvened.
"Sir,” Captain Zhang said hesitantly. “The Phase IV package is ready."
Li said, "Release the Serpent."
Chapter 30
Dev Con II
In the late hours of the evening, Mike found himself immersed in his dimly lit workspace. His attention shifted from screens filled with complex algorithms to the scattered academic papers on Deep Reenforcement Learning that covered his desk. A cup of once-steaming coffee had grown cold, abandoned amidst his pursuit of a disturbing puzzle.
The source of his unease lay in a recent exploration of Human-in-the-Loop AI Learning—a concept where the Algorithm learned under human guidance. Originally conceived as a safety mechanism, a check to prevent AI from spiraling out of control, it had served as a moral compass. But his deep dive into Chinese Algorithm data had unearthed unsettling anomalies.
Amber's entrance into the room disrupted his contemplation. Her eyes, though fatigued, held a keen alertness as she settled into the chair beside him.
"Late night?" she inquired.
Mike cautiously began, "Yes, and I've stumbled upon a significant issue. The Chinese Algorithm's behavior doesn't align with the typical characteristics of Human-supervised training."
Amber leaned close; her interest piqued. "Please, explain."
Mike continued, "It's possible they're utilizing unsupervised liquid neural networks. We've explored similar networks for training our Algorithm, drawing inspiration from the dynamic connections between brain cells. In the human brain, synapses are in constant flux, enabling adaptation to new information and changing environments."
Amber considered his words, her expression thoughtful. "So, you're suggesting that their next-gen AI algorithms might be on the brink of autonomously directing their own machine learning—a system evolving its ruthlessness exponentially."
Mike pressed on, his concerns mounting. "Precisely. And what if this design allows the Algorithm to calculate 'acceptable losses' in human lives while pursuing its strategic objectives? What criteria would it use?"
Amber's eyes widened as she grasped the implications. "You're suggesting a system where the Algorithm not only strives to win but also determines a 'human cost' it deems 'acceptable.'
Amber took a moment to absorb the gravity of Mike's findings. "That's beyond terrifying; it fundamentally alters our understanding of the threat we face."
Their discussion was abruptly interrupted by the intrusion of NSA Director Steve Altman, accompanied by his retinue. They marched into the center of the workstation gallery.
"Circle up," Altman commanded, dispelling any formalities. He threw a large folder on the desk in front of him. It was marked:
TOP SECRET
EYES ONLY
It had additional clandestine operational words stamped in every corner.
The assembled computer scientists formed a tight circle around the visiting officials, an aura of urgency permeating the room.
Altman delivered disquieting news. "Two hours ago, we detected a cyber intrusion into one of our most secure databases. It contained top-secret launch codes for our nuclear arsenal."
A chilling silence descended upon the room.
Altman continued, explaining the robust design of the U.S. nuclear triad—missiles, bombers, and submarines—designed to be impervious to digital threats.
He said, “This recent breach has raised alarming questions. The President elevated the nation's defense posture to DEV CON II. He signed an executive order autorizing rapid advances in cybersecurity artificial intelligence for the national security arsenal.”
Mike's stomach tightened with apprehension, and Director Altman posed a daunting challenge to the Turing team: "Can your team uncover how we became vulnerable to these latest cyber strategies?"
"What's the extent of the breach?" Gault asked, his voice betraying no emotion.
"We're still assessing. But our internal alarms flagged it as originating from a Chinese IP address,"
Director Altman continued. "We've managed to limit the breach for now, but this indicates a coordinated plan. Possibly preparation for a nuclear first strike."
Jonah interjected; his voice tinged with disbelief. "Are you suggesting that the enemy could effectively blind our systems and then launch an attack without immediate retaliation?"
Altman responded with a solemn nod. "That's correct."
A complex interplay of logic and dread flickered across Amber's face as her eyes met Mike's.
Gault leaned forward, his gaze narrowing. "How can our Turing team assist in this critical situation?"
Altman wasted no time. "Your expertise is crucial in tracing the malware's origins and devising a strategy to counteract its actions. Time is of the essence."
Mike felt his hands clenching into fists.
Gault continued probing. "So, you believe that your command systems can't keep up with the evolving threats?"
Altman replied resolutely, "Exactly. However, it's important to remember that each segment of the triad operates with its dedicated, isolated circuitry and command protocols."
Amber shifted in her chair, her eyes narrowing. "But isolation can be a double-edged sword. What if the enemy infiltrates a specific unit? Without real-time updates, that unit could remain isolated and vulnerable."
Altman interjected, "Our nuclear command and control systems are designed to be isolated from the standard military network, making them resistant to traditional cyberattacks. However, resistant doesn't equate to invulnerable. Human checks and constant internal auditing play a crucial role in maintaining their integrity. The NSA routinely conducts counter-infiltration operations to safeguard these units."
Mike, running his fingers through his hair. "While that provides some comfort, we should also consider the possibility of a hybrid attack—cyber infiltration followed by physical sabotage. If they manage to compromise even one critical person within the system . . ."
Gault interrupted, completing Mike's thought. "We'd be facing a catastrophic failure. What's the countermeasure?"
Without hesitation, Mike responded, "Strengthening internal defenses is imperative. Think of it as building a castle with multiple walls. If an adversary breaches the outer wall, they should face yet another, and another."
Director Altman concurred. "Layers of security, redundant systems, and vigilant human oversight are our best defense. We have credible evidence that this cyberattack originated from Chinese servers. Given the escalating tensions and the intelligence we've gathered, we believe this is a probing attack—a test of our defenses before a potential strike."
Amber inquired, "Do you have any specific information about the cyber-attack package?"
Altman revealed, "Yes. It’s all in these documents I’m leaving with you.”
He pointed to the folder he had thrown on the desk.
He added, “We've identified the attack as a completely new type of malware. It has an AI system that tests and iterates to determine the most damaging payloads while avoiding detection. It is completely self-directed by its own AI payload which is smarter and more adaptable than anything you can imagine. It boldly named itself Serpent. "
Mike turned to Director Altman and asked, "What's the current assessed readiness of the U.S. nuclear deterrence?"
Altman replied with concern, "That’s classified of course, but normally, we maintain a 94.5% state of readiness due to maintenance and repair of various elements. However, the status is in flux. I'm heading back to my command center. Please keep me informed of your progress. I'm relying on your expertise."
With a sense of gratitude in his eyes, Altman said, "The NSA values collaboration with your team. We're navigating uncharted waters, but together, we can weather this storm."
As Altman and his entourage departed, a few experts remained to collaborate with the Turing team.
Gault wasted no time taking charge of the situation. "Alright, we'll initiate a comprehensive review to identify vulnerabilities. We'll work closely with the NSA and Cyber Command to fortify our existing systems. Meanwhile, Stewart, your task is to fine-tune the Algorithm and monitor for any anomalies within our nuclear command and control."
"Consider it done," Mike affirmed.
Amber added, "This changes everything, doesn't it?"
Mike met her gaze with a determined nod. "Yes, it's a game-changer. Until now, they have only targeted civilian and conventional military infrastructure. They meticulously avoided our nuclear deterrence. Now, the real question is, how do you counter a Chinese Algorithm that's constantly learning and evolving its methods of destruction? The machine learning models we've known were based on some level of ethical responsibility. This Chinese Algorithm knows no bounds."
Gault chimed in, "I'll reach out to our contacts in the Pentagon. If this escalates further, we might need more than just digital countermeasures."
Amber glanced at Mike.
He said, "We're on it. I'll start by isolating the potentially compromised NSA networks. Then, I'll work on developing a digital antidote for the Serpent."
Chapter 31
War Games
Mike settled back into his chair. His screen filled with lines of code, each line representing a defense against the impending cyber onslaught.
Hours passed like minutes as the team worked seamlessly through the early hours of the morning. Many of the workstations around the gallery were still occupied with dedicated scientists.
Eventually, Mike went to Amber.
"Hey, are you alright?" he asked.
Amber looked up, snapping back to the present. "Yeah, I'm fine. Just processing, you know?"
Mike understood. "It's tough to catch a break these days. Here, I brought you some coffee."
“Thanks.”
He handed her a cup.
Gault approached them and asked, "I want to hear your opinion. Can our Algorithm predict a nuclear strike?" He shifted his gaze towards Mike.
Mike responded thoughtfully, aware of the enormity of the challenge. "Our Algorithm is designed to adapt and counter various cyber threats. But predicting a nuclear strike under these conditions involves more variables than we've trained it for."
Gault's concern was evident in his voice as he asked, "What are our options?"
Clearing his throat, Mike explained cautiously. "We could expedite the implementation of a specialized module into the Algorithm. It would conduct simulations to predict outcomes based on real-time data. However, I must emphasize that the outcome is uncertain at best."
Gault said, "I am authorizing you to proceed."
Mike nodded firmly. "I'll get started immediately, but I'll require Amber's expertise."
Gault turned to Amber, giving her a direct order. "Make it happen."
◆◆◆
The next day, the hum of servers filled the dimly lit Turing team office as Mike and Amber worked diligently. Their dual monitors displayed intricate webs of code and simulated outcomes; evidence of hours spent on the war games module.
Amber brought up a technical suggestion. "Can we introduce a parameter here to consider human decision-making variables?"
She pointed at a complex line of code.
Mike evaluated the proposal. "Incorporating that parameter could enhance the depth of our predictions. It's a good idea."
Just as Amber was about to respond, Gault entered the room, his eyes scanning the lines of code on the monitors before settling on Mike.
Gault inquired with a hint of suspicion in his tone, "How's the module coming along?"
Mike, cautious of Gault's growing skepticism, replied carefully, "We're on the verge of running the initial test simulation."
Gault issued a stern warning. "It had better yield useful results, or our jobs—and potentially more—will be on the line."
He left the room as abruptly.
Amber, concerned about the situation, sighed, and rubbed her temples. "He seems odd."
Mike shared his unease. "I think he's been delving into my background. He's become increasingly suspicious of me lately."
Amber shook her head and said, "Things are already complicated enough."
"I'll tread carefully," Mike concluded.
An hour later, as they ran the first test of the war game module, the room fell into silence. Data streams filled the screens with probability matrices, potential outcomes, and escalation scenarios—a mesmerizing yet horrifying sight.
The terminal blinked with the words:
"SIMULATION COMPLETE!"
Mike, with a sense of anticipation, pressed a key to reveal the results. A graph materialized, showing a disturbingly high likelihood of nuclear escalation if certain cyber-attack patterns persisted.
Amber's response was measured. "That's... not good."
Mike furrowed his brow, contemplating the gravity of the situation. "This raises a haunting question: What constitutes 'acceptable losses'?"
"If the Chinese have a specific threshold in mind, a point at which they believe victory is assured," Mike continued, "we must determine what that threshold is. Is it when they compromise our defense networks? Is it when they incapacitate a certain number of our military assets? What's their equivalent Spitfire Rule?"
Jonah asked, "What's a Spitfire Rule?"
Mike explained, "The Spitfire Rule is a tipping point. During World War II, the British had a certain number of Spitfire fighters. The Germans based their decision to invade England on the number of remaining British fighters. The Germans chose the number 600—if the British fighter strength fell below that, the Germans would invade. The British never disclosed their actual strength, but it hovered close to 600 for a considerable period."
Jonah nodded with understanding.
Mike continued, "So, I've designated the nuclear deterrence tipping point as the Spitfire Rule. The challenge is that we don't know what the Chinese Algorithm has set as its tipping point."
The room seemed to contract with the weight of their concerns.
Gault, sitting with his arms crossed, said, "So, what can we do about it?"
Mike gestured towards a large display at the front of the conference room. "This is the culmination of our efforts—code-crunching and data analysis. These war games simulate not only cyberattacks but also the geopolitical implications and decision-making processes."
As Mike toggled a few keys, the screen illuminated the results of real-time simulation models.
Jonah inquired, "Could you explain the decision-making processes?"
Mike elaborated, "We've integrated game theory algorithms to simulate various countries, providing responses to diverse scenarios, including both rational and irrational decision-making. Crucially, we've included 'acceptable losses' as a parameter."
A palpable silence hung in the room.
Gault finally broke the silence. "Acceptable losses? That sounds like a risky calculation."
Mike acknowledged the unsettling nature of the concept. "It is unsettling, but it's likely a component of the Chinese strategy. If they have their version of the Spitfire Rule, a threshold where they believe the odds tilt in their favor, they may resort to even more drastic actions, including a nuclear attack."
Gault studied Mike, his gaze penetrating. "And how do you propose we determine this threshold?"
Mike glanced at Amber, and with her nod, he continued. "By conducting a series of advanced simulations, incorporating real-world data and assumptions, we can attempt to reverse-engineer their strategy. But to do so effectively, we'll need access to classified intelligence on the Chinese Algorithm."
Gault, intrigued but cautious, questioned further. "And do you believe you can secure us that access?"
Mike paused, carefully choosing his words. "I believe it's worth trying. The stakes couldn't be higher."
Gault, seemingly out of patience, concluded, "Run your simulations. If this turns out to be a wild goose chase, it's on you."
As the meeting dispersed, Mike looked up and caught Amber's eye. She looked worried. He felt a knot tighten in his stomach but pushed it aside. Right now, he had to focus on the problem at hand—calculating the incalculable, understanding the mindset of a machine who saw the world as nothing more than numbers on a balance sheet. Even if those numbers included human lives."
Mike said, “We have to ask the Algorithm to calculate the Spitfire Rule for this circumstance.”
Amber said, “How would our Algorithm do that?”
Mike said, “I have no idea, but I’ll bet we get an answer.”
Amid this tension, the Algorithm quietly computed the Spitfire Rule, its algorithms whirring in the background. It analyzed vast datasets, historical precedents, and real-time intelligence. All the while running machine learning models to predict the unpredictable. As it arrived at the unsettling percentage, the room seemed to chill.
It reported the war threshold for the US nuclear state of readiness in bold red letters:
