Chasm Waxing: A Startup, Cyber-Thriller

  “That’s all you need to know for now. But that’s not all I want,” said the General. He poured them both another scotch. “I want you—now.”

  Shields’ words didn’t register with Samantha. “Why don’t you like Becca?”

  “I like Becca a lot. I agree. She’s one of your best. I know Saul is a better coder, but Becca is both a great software engineer and an outstanding communicator. She’s probably more valuable to Gamification Systems than Saul. You can put her in front of a client. It’s great that you have her working with Josh. But I’ve read her SF-86.”

  SF-86’s were the form that individuals seeking to work with classified information submitted to the government. It was the document used to formally request a clearance. The Office of Personnel and Management—OPM—was the government agency ultimately in charge of SF-86’s.

  SF-86’s not only contained extremely personal information, like financial records and health data, but also included data on known associates of the person seeking the clearance. This included neighbors, co-workers, parents, spouse, ex-spouses, and children. In 2015, the OPM database was hacked, yielding a massive amount of personal information—including fingerprint data.

  The General continued, “Her dad is a fundamentalist pastor. She strikes me as a bit naïve. Becca is a straight arrow. Straight arrows get bent when you have to fight radical Islamic terrorists, like the Caliphate.”

  A flash of hot anger shot through Samantha. Who the hell was he, to tell me, how to run Gamification Systems? She quickly suppressed the emotion.

  Then, the General’s words of boardroom foreplay finally hit her. She was appreciative of him for the Gecko opportunity and funding offer. She didn’t want to ruin the moment. “Ok. I understand.” She’d employ the art of persuasion in upcoming days.

  Samantha rose from her chair and straddled the General with her taut thighs.

  As the two began to kiss, the General fondled Samantha’s voluptuousness.

  “General,” she said, oozing lust, “I don’t know anything about a straight arrow; but, there’s something else that’s very straight in this room.”

  Samantha grabbed the General’s hand. She led him over to his English oak desk. In one fell swoop, Samantha dumped the entire contents of the desk to the floor. Papers and file folders scattered everywhere. Then, she sprawled on the desk and beckoned the General to join her.

  The two pleasured each other with passionate sex for hours.

  Chapter 11 – Becca Meets Josh

  9:05 a.m. (EDT), Tuesday, July 28, 2020 – Columbia, MD

  Suite 201, Gamification Systems’ Offices, Defense Innovations Accelerator

  Josh exited the elevator and set foot into the offices of Gamification Systems, for the first time. He smiled at the receptionist, asking if Becca Roberts was available. “My name is—”

  “Josh Adler,” interrupted the receptionist, reciprocating Josh’s smile. “Let me see if Becca is available.”

  Becca said she’d be right out. She reached into her purse, grabbed a mirror, and fixed her shoulder-length hair. There was no ponytail today. Eye makeup replaced her glasses. Like the receptionist, Becca knew of Josh Adler.

  Over the past nine months, they had on occasion, seen one another in the Accelerator’s cafeteria. They exchanged niceties—but nothing more than that. They were competitors.

  Becca entered the reception area. “Hi, Josh!”

  “Hey, Becca. Thanks for meeting me on such short notice.”

  “Sure, what the General wants; the General gets. Let’s go back to my office to talk about integration options. Do you need anything to drink?”

  “I’ll have a shot of tequila,” he answered, sarcastically.

  Becca laughed. “So early in the morning? Yesterday’s demo went that well, huh? How about coffee? Then, let’s see if we’re still talking by happy hour.”

  “Deal.”

  After stopping by Gamification’s kitchen, Becca walked Josh back to her glass office. All the see-through offices overlooked a room that contained a conference table and wall-to-wall whiteboards.

  “Cool offices,” said Josh. “Wouldn’t I love to see what my employees were up to, every minute of the day?” Next door, Samantha was on the phone with her door shut. She waved at Josh. Becca’s workspace was manageably messy, with lots of technology books and magazines strewn about. She must have owned every edition of 2600 Magazine.

  Josh examined Becca’s degree in computer science. “How’d you like Carnegie Mellon?”

  “It was an excellent education, but I hated the Pittsburgh weather and stinkbugs. I wish Andrew Carnegie had started the steel industry by a beach.”

  “Ha!” bellowed Josh. Next, he focused on a picture of a much younger Becca. She stood, arm-in-arm, with a man that looked to be her father. They were dressed in fall season, camouflaged hunting gear. A black boar lay dead at the tip of their hunting boots. He’s kind of nosy, thought Becca.

  “You’re a hunter?” asked Josh, with a surprised inflection.

  “Yeah. My dad taught me to hunt, while growing up in Texas. That picture was taken at a family friend’s ranch. We’d go there to kill wild pigs. Farrell hogs are nasty. They’re all over Texas ranches. If you don’t kill them, they’ll eat all the crops and make huge mud pits in the fields. They love corn.”

  “I’m impressed. You’ve got to be brilliant to get into Carnegie’s computer science program. And you can handle a gun.”

  Becca giggled. She hoped she wasn’t blushing.

  Josh grabbed the chair in front of Becca’s desk and got down to business. “During my demo, General Shields suggested that I meet you, to see if we could help each other with AI. AI is the heart of my company. I’ve hit a roadblock in my machine learning and NLP algorithms. Right now, our AI is narrowly applied to recognizing cybersecurity events—scans, botnet attacks, deployment of malware—things like that.

  “CyberAI also matches these cyber-events against threats we gather from crawling the web and social media. So for example, if a hacker starts maligning your company in a Tweet, we’ll be on elevated alert. The General thought we might be able to help you recognize these cyber-events. He also suggested that you had a more general purpose need for AI. I’m not sure what he meant by that.”

  Becca gazed into Josh’s face—and dimples—intently. Luckily, her multi-tasking mind could listen at the same time. “I think there are a couple of possibilities for integration.”

  She picked up a coffee mug full of whiteboard markers and spied an empty section of her office wall. On the glass, she drew the same architectural diagram she sketched for the General—REALSPACE, G-Bridge, and GAMESPACE.

  “We turn cybersecurity into a game. But that’s just one use case. We want to be able to gamify a lot of things. So, I see some synergies here. Quite frankly, we’ve spent a lot of time programming our system to recognize cyber-attacks and events. If you could do that for us, it would allow us to focus on G-Bridge. G-Bridge has a robust API. So we might be able to rip out our AI, and replace it with yours. What language did you use?”

  “The architecture of the CyberAI engine is pretty sophisticated. It’s got packages written in C++, Java, CUDA, Python, and even some Assembly Language. Some modules have to run like lightning. But the CyberAI engine exposes a number of APIs, so to you, it’s like a black box. We have APIs for C++, Java, Python, and REST.”

  C++, Java, CUDA, Python, and Assembly Language were all computer programming languages. Each language had strengths and weaknesses. CUDA was a language that you’d use to interact with NVIDIA Graphics Processing Units—GPUs. GPUs were special purpose chips that offloaded certain types of tasks from a computer's CPU. The first market for GPUs was gaming machines. Recently, car and truck manufacturers were rapidly integrating GPUs into their autonomous vehicles for their AI incorporating deep learning algorithms. GPUs were also critical c
omponents within VR and AR gear.

  Assembly Language was a low-level language that talked directly to CPUs. It was complicated to write, but ran very quickly. C++, Java, and Python were all higher level languages. They were easier to write than Assembly, but executed more slowly. The REST API spoke the same language as an Internet browser. Companies, like Facebook and Twitter, exposed REST APIs to enable computer programs to interact with their websites—without the use of a browser.

  Josh continued, “I’d recommend either the C++ API or the Java API. Those will work best for you.”

  Becca was encouraged. “That’s awesome! Are you guys able to handle spear phishing?”

  “Yes, sometimes we throw a false positive. But at least you’re made aware of the danger. The spear phishing engine detects spoofing and homographs. It also incorporates DNS heuristics and can see if emails comply with an enterprise’s content compliance policies.”

  “Great,” replied Becca. “If you’re better than us at recognizing cyber-events in REALSPACE, especially things like spear phishing, then we’ll just use your AI engine. I think the second place will be more of a stretch for you. It’s beyond the weak AI of merely recognizing cyber-events. It’s related to GAMESPACE.

  “I’d like to investigate how your AI could replace or, at least lessen the need for our G-Master. This requirement should help you move your engine towards stronger AI, because you’d be replacing a human being.

  “Right now, a G-Master has to watch every moment, of every game. The G-Master has to ensure the Gamers don’t get stuck, and that the cyber-events in REALSPACE are appropriately represented in GAMESPACE. For example, what if the Gamers collectively decided to seek treasure instead of thwarting a virus in REALSPACE? Instead of fighting a monster in GAMESPACE, they’d be filling their pockets with gold.”

  “Who could blame them?” said Josh, sardonically.

  “That’s the problem. In our game we call Castle Gecko…I mean Castle Chevaliers, we try to handle this with Bitcoin rewards. But, Gamers aren’t forced to do anything. It’s an open-world game. As G-Master, I might have to intervene by spawning a non-player character or verbally chastising them to get back into the fight.”

  “I see,” said Josh, “so your client is Gecko Insurance?”

  Becca went flush. “No, our client builds castles,” she replied, with a deadpan voice. “The point is, I could see using your AI to replace our G-Master.”

  “I understand. I agree that we can definitely take over the AI for REALSPACE when you’re dealing with cybersecurity. And we can work further on your G-Master replacement requirement. It does fit with my vision of developing stronger artificial intelligence. Let me tell you a little more about CyberAI.”

  Josh stood up and moved towards Becca. “May I?”

  “Sure.” Becca handed Josh her coffee mug full of whiteboard markers, and the red pen in her grasp. In the exchange, Becca and Josh’s hand touched—just for an instant. Becca’s stomach tingled. All she could see were dimples. She thought their touch might have passed a static electric charge.

  Becca struggled to get back to her seat without drawing attention to herself. “You can erase all of that.” Phew, I’m glad my voice didn’t squeak, she thought.

  Josh penned his own architectural diagram. “It sounds like you’re very familiar with AI. CyberAI is a comprehensive cybersecurity suite that touches network security, application security, server security, monitoring, smart availability throttling, and automated incident response.”

  Becca laughed. “How in the world do you remember everything you do?”

  Josh grinned. “I don’t. The AI does. Anyway, one of my special focus areas was mitigating insider threats—rogue system administrators and hackers that game the system to get elevated privileges. If a system administrator—like Edward Snowden—starts reviewing a lot of files or searching in directories where he’s never been; CyberAI raises alarms.

  “I’m trying to enhance the AI. I want it to predict when someone will be a threat. The General really likes that part. And of course, we integrate with virus and malware detection software. I’ve already told you about how we ingest data from the web and social media sites. I’m also trying to create an AI-bot that crawls the Deep Web. So that’s a broad overview.” The Deep Web was the portion of the Internet that was not indexed by search engines.

  “Our AI Kernel is at the heart of all of this. We developed it to work with various AI algorithms. Until last month, the AI Kernel focused on text—you know with NLP—and machine learning. The algorithms operated on both the log files and the web data. Of course, the perfect NLP would read a book, and understand it as well as a human reader. My NLP is far from perfect.”

  Josh told Becca about his MIT coursework and the genesis of CyberAI.

  “I used one-third of the security log data to train my NLP machine learning algorithms. Then, I used the other two-thirds of the data to see if I could detect the attacks. I repeated this process of training and testing, training and testing, until I was achieving good results. My CTO, Vish Kumar, focused—”

  “Vish Kumar from Graphica Intelligence? He’s a rock star!”

  “Yeah,” replied Josh, “He came to us after Graphica’s acquisition. Vish’s team is superb at packet inspection and employing graph analytic algorithms to identify threats. His stuff is slick. It proactively identifies suspicious packets from compromised hosts.”

  Graph analysis used graph theory to find meaningful relationships in social networks and other graph data. Graph analysis of your Facebook friends or Twitter followers could lead to discovering unexpected connections and relationships.

  “So why are you in my office?” said Becca, semi-sarcastically. “It sounds like you got this all figured out.”

  Josh huffed. “Hardly, CyberAI only recognizes 83% of cyber-threats. That’s too low. It doesn’t do enough to minimize the need for human security administrators. Now Vish’s stuff is killing it, but my software isn’t improving in its recognition capabilities. And that’s just recognition. I’m not moving any closer to where I really want CyberAI to go—prediction. And beyond that, I want to enable discovery. I think the future of AI is extending human intelligence, not replacing it.

  “I’ve gotten the last drop of blood from improvements to NLP using standard machine learning techniques. Last month, I decided to add another group of algorithms to our AI kernel—deep learning. Do you know anything about deep learning?”

  “Yes, but act like I don’t,” she replied, channeling her inner General Shields. Becca did draw a line. She wouldn’t instruct Josh to pretend she was his grandma.

  “Deep learning is a particular type of machine learning. Deep learning algorithms break problems up into many different layers. Statistical calculations are performed on the data at each layer. The information can be images, sound, text, graphs, and so on. These calculations determine the essential similarities, or features, for each layer.

  “Deep learning is inspired by learning in the human brain. The multiple layers of a deep learning algorithm are called a, ‘neural network.’ Powerful neural networks, trained by deep learning, are behind some of the biggest breakthroughs in AI in the last decade.”

  “Breakthroughs like what?” asked Becca.

  “Breakthroughs like self-driving cars. Autonomous vehicles must recognize and react to obstacles and road conditions. Breakthroughs like speech recognition in digital personal assistants. Apple’s Siri, Microsoft’s Cortana, and Amazon’s Echo have to understand speech to answer questions and process commands. Breakthroughs like Facebook’s software that automatically tags and categorizes images without human help.”

  “That’s interesting. I didn’t realize that deep learning was so pervasive.”

  “Who knew, right?” said Josh. “Deep learning is becoming a new computing model. It’s going to impact every industry. Not only is it going to a
llow voice commands to be a new input for computing, but visual computing will take off. Kids will play chess against the computer with a real chess board.

  “Despite all its promise, I’m struggling to get any meaningful impact from my neural network that understands English. I’ve crawled the Internet with spiders, just like the Atom search engine. I’ve assembled an enormous amount of text to use as a training set. While my hardware certainly can’t compare with Nucleus’ server farm, my dad was a big investor in NVIDIA. He gave me two, first-generation NVIDIA DGX-1 deep learning supercomputers.

  “I used this data to train the neural network to understand the text. Then, I applied the neural network to the log files I’ve used in the past. When I combined my old machine learning stuff, with new deep learning algorithms; the AI’s inference results only improved .08%. That’s why my demo for General Shields went so poorly.”

  Becca carefully examined CyberAI’s architecture. After some long moments in thought, she said, “I believe you need a better training set. You need a purer corpus of text that serves as your ground truth. My guess is that there are too many semantic differences in your Internet text. Your neural network isn’t understanding the English well enough.”

  Josh looked impressed. “I’ve got a pretty sophisticated annotation layer that allows me to add nuances and labels to the corpora. But I agree; that’s where I need to focus.”

  “Why not use the Bible?” suggested Becca. “The Bible has a vast amount of text and many different versions. They all carry the same ideas using different words. That can help you with the semantics. If you ever want to move beyond English, there’s a Bible for every language. Plus, you have a massive amount of commentaries written in different time periods. There’s a shared meaning between all of this corpora to help you with semantics and labels. It’s like a built-in annotation layer. It’ll give you many more reliable hooks.”

  With a droll smile, Josh replied, “The Bible? Are you some kind of Holy Roller? Hacker, computer programmer, hunter, and Bible thumper?”

  Becca scowled at Josh.

  “What? I’m just having fun.”