The Hive

by Barry Lyga

  “They got all of these guys together. Not one of them had all of the pieces of the puzzle, but together they did. So they glued them all together behind the scenes. Get some data from Apple, some data from Google, some more data from Facebook … They took advantage of the Spectre and Meltdown vulnerabilities before they were made public. All these flaws, just waiting to be used. You mash it all together in an NSA supercomputer somewhere, and it spits out BLINQ. One ring to rule them all.”

  “Come on,” Cassie scoffed. “Those are some of the most powerful people in the world, running the most powerful companies in the world. You can’t just make them do things, even if you’re the government. Hell, I remember my dad telling me Apple wouldn’t let the FBI break iPhone encryption. You think they’re just going to go along with the president?”

  “You think they had a choice? They hide behind the law,” Carson told her with disgust. “ ‘We just follow the laws of the countries we do business in,’ they say.” He tapped some keys and a web page came up. Cassie skimmed it.

  “Section 702 of the FISA Amendments Act,” Carson announced. “Between that and Section 215 of the Patriot Act, they modified the legislation that had originally set up a secret court called the FISA court. Made it easier for the government to spy on people. No one knows what goes on in there, but according to legal blogs I’ve checked, the Justice Department could totally use it to force concessions from the tech companies.”

  “The Justice Department …” They administered the Hive.

  “Yeah. They used FISA courts to bend the tech companies to their will in secret. No one knew. No one could tell.”

  “So the government forces all of the companies to cooperate, to use their data, their algorithms and their encryption to create BLINQ,” Cassie said. “Was Alexandra lying? Was it really not about policing people online?”

  “It’s a twofer,” he said, shrugging. “Misdirection, like a magic trick: ‘Hey, look over here at the shiny new internet thing that lets you shame your neighbors!’ And while people were doing that, this administration was laying the groundwork to use the same system to control people.”

  Cassie swallowed hard. “By letting them think they were in control. But they weren’t.” She frowned. “We weren’t,” she admitted. She’d been as big a Hive booster as anyone, until it was pointed at her. The whole thing was superseductive and it had worked like a charm.

  “But why pick me?” she wondered. “And how did they do it?” She thought about what Alexandra had told her. “Alexandra said that I was a convenient scapegoat.”

  “Well, yeah, sure. Misdirection again. What they really wanted was a way to test Level 6, right? To make sure people wouldn’t rebel against it, to see if someone would actually kill someone just because of what she or he said online.”

  “It didn’t matter that it was me,” she murmured.

  “They were looking for something, anything.”

  “Convenient.”

  Carson leaned back. “Yep.”

  Cassie thought back to what Alexandra had told her and Bryce at Venecia. “So they used these things called ghost accounts,” she began. She filled him in on the rest of what Alexandra had told her.

  He sighed when she was done. “This is what I don’t get. I don’t care what Alexandra told you: you can’t game this system. It’s impossible.”

  Which was true. Every participant (that is to say, every American citizen over the age of thirteen) was Verified and given a unique twenty-two-digit identifier. With such a long identifier, there were trillions of possibilities, but only a small percentage were valid.

  It was the same principle used by credit card numbers. Credit card numbers were superlong not for identification purposes but rather for security. If you tried to create a random credit card by stringing together the right number of digits, the odds were overwhelming that you’d “hit” one of the many, many trillions of numbers that weren’t valid. And with so many trillions of possibilities, your odds of luckily hitting a “good” card number were abysmally low.

  The same logic ran BLINQ identifiers. With so many numbers and so few (relatively) valid ones, it was impossible to create bots or false accounts.

  “You’d fail the checksum every time,” Carson said. “The system would reject the identifier as invalid and not create the account. So, no bots.”

  “But Alexandra said there were ghost accounts. She said that’s how they’ve been ramping up my Condemns. Are you sure you can’t fake the identifiers?”

  Carson gestured to one of the screens, which scrolled endlessly with code. Cassie had ignored it before because it was just flowing by so fast, but now she forced herself to scrutinize it.

  It was an endless stream of twenty-two-digit numbers, followed by the words “Processing” and then “Invalid!”

  “TonyStark put this together, but I’ve been running it for him because my bandwidth is better,” he told her. “It’s a megacluster with a so-called dark matter IP to generate billions of potential BLINQ accounts. They’re all ditched by the system as invalid. None of them pass the checksum.”

  Cassie let the scrolling text hypnotize her. Sometimes if you let the code bleed away, answers revealed themselves. Sometimes if you just stared at everything and nothing at once, the world melted. In the thaw you found things you didn’t even know you were looking for.

  “Who decides what’s a valid identifier in the first place?” she heard herself ask. She was on a path. Not sure where it went, but it was something.

  “Well … BLINQ does,” Carson told her. “They assign it when you turn thirteen and your parents file your Social Media Induction form.”

  “Right.” The path was looking a little clearer. “And they generate the number the same way credit card companies do, right? They assign it and it goes into a database somewhere.”

  “Sure.” Carson shrugged. “And then it works just like when you charge a Starbucks or a ride share. You plug in a number and the system makes sure it matches in the database.”

  “So what if BLINQ has made and approved a whole batch of numbers that don’t actually connect to anyone? What if only they know about these accounts? And what if they’re using them to game the algorithm and direct Hive Justice where they want it to go?”

  Carson shook his head. “No. Doesn’t work. Ever since 2016, everyone knows what bots are like. I’ve been examining your virality and the interactivity of your hashtags. Assume for the sake of argument that someone has figured out how to make BLINQ bots. There’s nothing in the activity around you that looks like a bot. There are characteristics that bots have and we know what they are, and I’m not seeing them.”

  Cassie frowned. Bots were handy, yes, but limited and they could only do so much. Made for a specific purpose, they stuck to that purpose. If you wanted to persuade people to eat vanilla ice cream, you could have a bot tweet at someone who talked about, say, chocolate ice cream and spam them with vanilla ice cream coupons and information. You could have the bot interact with people who said things about vanilla ice cream to reinforce their behavior. But no matter how good the bot, it wouldn’t pass the Turing test because all it would care about would be vanilla ice cream. And pretty much any decent hacker could identify that in five minutes.

  “OK,” she admitted, “so if it’s not bots, then what? Who?”

  “Throw in how, when and why,” he told her.

  “We know some of those. But, yeah, this whole thing feels like trying to unravel a spiderweb without breaking any strands.”

  A thought occurred to her. She jumped up and commandeered Carson’s chair. There, on the center screen, was her BLINQ account. Likes: 90 234 491. Condemns: 140 384 889. She grabbed the mouse, then found herself tapping it rapidly on the desk, unsure what to do next.

  Carson came up beside her and very kindly did not chastise her for stealing his chair and touching his gear.
>
  “What are you thinking?”

  “Is there a way to see who hasn’t voted?” she asked. The thought formed as she spoke the words. “Apathy,” she’d said, and Alexandra had agreed. The original plan for the Hive had anticipated a counterbalance to the system’s worst impulses, but they hadn’t accounted for apathy, for some people not caring enough to click. “Can we get a count of people who haven’t bothered?”

  “Abstentions?” Carson slid into the guest chair and leaned forward, chin on fists. “That’s not a public-facing API, but we know there are about 350 million BLINQ accounts, so you just subtract your Likes and Condemns … It’s like a hundred million or so.”

  “No.” She clicked around, opening a Terminal window. An API was an application programming interface — it was the way programmers communicated with code they hadn’t written. Systems had APIs that hooked into aspects of their functionality — if there was an API that, say, controlled a system’s fonts, then you could write code that used that API to manipulate text in your own app without having to recreate it all from scratch. Each company decided which public-facing APIs would be made available to any programmer.

  But there was another way.

  “Can we get to the private API?” she asked. “They have to have something like that set up for internal use, right? They’d need it so that they know how many ghosts to use. We need to do this right.”

  Carson nudged her chair with his foot, pushing her out of the way. “Let me see what I can do.”

  He put the word out on the OHM network. There were still groups of them, of course, all around the country. In the wake of the raid on TonyStark’s crew, most of the cells had gone dark in self-defense, but Carson was able to track down a couple of loners who had the information they needed — the IP addresses associated with a BLINQ server farm on the Canadian border.

  He leaned back, frustrated. It was one thing to have the IP addresses — that was like knowing where someone’s house was. Beyond that, though …

  “I know it just looks like a bunch of code on the screen …” Carson told her and drifted off.

  She’d done enough digital breaking and entering in her time to know. They knew where the house was, but it was surrounded by a wall sixteen feet thick and ten miles high, and there was a bloodthirsty pack of wolves roaming the front lawn.

  “BLINQ security is insane,” he said.

  Cassie pursed her lips and took over. She drummed her fingers on the desk, thinking.

  She poked around the edges of the facility’s systems. Nothing that would trigger any alarms. Just getting the lay of the cyber-land.

  “I think I see something that’ll work,” she said after a little while. “Can you put together a database query while I get us through security?”

  He blinked. “Are you serious? You’re gonna hack a government system? Just like that?”

  “Just like that, yeah.” She flexed her fingers and started typing.

  By the time Carson had crafted the query they needed — set up to mesh with the specifications of the BLINQ API and its specific database structure — Cassie had evaded BLINQ security and managed to open a port into the database. The system was nearly impregnable, but once again human fallibility met brute code and spat out garbage. A sysadmin had dutifully updated the security system to counter the KRACK II vulnerability, but had ignored several smartbulbs. They’d never had their firmware updated, so Cassie was able to turn them into zombies to use as attack vectors to worm her way into the network. It was no easy feat, but it was totally doable. The Internet of Things was a real boon to hackers — once people installed their smart-stuff, they tended to forget about it and neglect the firmware security updates. Meaning it was ripe for exploitation by people like Cassie.

  Carson offered up a low whistle as he leaned over her shoulder, watching. “I feel like a Little League kid at the World Series,” he admitted.

  Cassie smiled. In so many ways, he reminded her of her dad … except Harlon had no humility at all. She grabbed the query from him and ran it against the BLINQ database. It wasn’t a complicated query — just scooping up any accounts that hadn’t voted on Cassie and sending the sum back as an integer — so it took almost no time at all. Nonetheless, Cassie killed the connection the instant she could, and then they sat frozen and staring at the screen for long seconds after.

  They had just hacked a secure government database without really thinking it through. They sat in silent, mutual awe. Had she gone too far? She thought she’d covered her tracks, but there was always a better system, smarter code, a more devious hacker. What kind of trouble might she be leading to Carson’s door?

  Carson broke the almost eerie quiet. “Uh, there’s a chance,” he said with a little shiver in his voice, “that my parents are going to get audited this year.”

  Cassie couldn’t help it; she burst out laughing. After a moment, Carson joined her.

  *

  The number, when they looked at it, was bigger than she’d anticipated. Well over two hundred million. Given what she’d been through in the past week, Cassie couldn’t imagine that more than two hundred million people hadn’t bothered to chime in on #AbortionJoke and #HasCassieSurfacedYet?

  And the fact was: they hadn’t. There couldn’t be that many abstentions. It was impossible.

  “Bots,” Cassie said triumphantly, folding her arms over her chest.

  Carson popped the tab on a can of CaffBomb!, the preferred beverage of those who can’t be bothered with sleep. “No BLINQ bots. Seriously.”

  “There have to be,” she said, and pointed at the number again.

  He peered at it blearily and swigged his drink. “So a bunch of people abstained. So what? You’re the social media devil of the day, but not everyone gives a shit.”

  Cassie swiped the can from him and took a long, generous swallow. “Let’s do some hard-core science here. Open up the calculator.”

  “Calculator? I do math with the command line.”

  “Of course you do. But let’s pretend we’re stupid.”

  Carson opened the calculator app. Cassie read off a string of numbers to him: her Likes, her Condemns and the abstentions figure from BLINQ.

  The total was significantly north of four hundred million.

  “Ta-da!” Cassie said.

  “So what?” Carson asked with a tinge of misery. “We did addition. Big whoop.”

  She took the can again, this time tilting it at his lips. “You need more caffeine. Your brain isn’t working right. More than four hundred million.”

  “Population of the United States is right around that …” He drifted off. “Oh, crap.”

  She nodded triumphantly. There were around four hundred million people in the U.S., but not all of them were over the age of thirteen. Not all of them were even online.

  “Ghost accounts,” Carson murmured, leaning in toward the screen, as if by getting closer to the data, he could somehow force it to make more sense. “Holy fucking shit,” he swore softly. “All of us too-smart hacker types were trying calculus and trig and computer science, and the answer was just plain old arithmetic all along.”

  “The ghost accounts exist. Alexandra wasn’t lying. The numbers are so big that they must have figured no one would bother to add them up,” Cassie said. “They hid it all in plain sight.”

  Carson snorted in derision. “You give them too much credit. I think they just went overboard. They got cocky and just kept juicing your numbers to make things worse and worse.” He snapped his fingers. “Plus, this is the first time they’ve ever done anything this big, with this many accounts. No one ever bothered adding them up before because there was no point — the numbers were so small that it didn’t matter.”

  “But this time, yeah, they went overboard. But then … how do they shield the bots from discovery in the first place? We know how bots work — they’re
easy to identify, but no one has been able to —”

  Carson suddenly jumped up from his chair and Cassie fell silent as he paced the room fiercely, knocking over piles of thumb drives, pawing through old Blu-rays.

  “Carson?” She didn’t want to interrupt this process, whatever it was, but time was not their friend.

  “Aha!” he crowed, and waved a USB drive in the air triumphantly. “Got it!”

  “Got what?” She joined him at the desk again as he jammed the drive into a port.

  “Patterns! I knew I remembered this, but I couldn’t remember where the backup was … Patterns! Mapping data!”

  “Mapping? What maps?”

  “Not like map maps,” he said, punching up a file. “Like, metaphorical maps. Maps of activity.” He pointed at a graphic filling the screen. “Here!”

  It was a complicated skein of digital silk, thousands of strands spun out from the center of a web, threads of multiple colors unspooling out to the boundaries of the screen and beyond.

  “What am I looking at?” Cassie asked, leaning in intently.

  Carson’s left foot tapped madly. “A map of BLINQ account activity for one of the accounts that was a big viral push for #HasCassieSurfacedYet? The threads are connections to other accounts, interactions, color coded.” He pointed out specific lines. “Likes, Dislikes, Condemns, retweets, BLINQ-ups. You get the picture.”

  It was a visualization of the account’s activity, she realized, rendering hundreds of interactions in a single image. “What time period does this take place over?” Bots were known for bursts of activity over short periods of time, targeted to do the most damage.

  “This is, like, a week,” Carson said, still tapping madly away at the floor. “Not typical bot behavior at all, right?”

  She shook her head. This looked, well, pretty normal. An opinion Carson then ratified by bringing up his own BLINQ activity graph. It wasn’t identical to the first one — that would have been weird! — but they had similar characteristics. For example, they both interacted with other accounts on a regular basis, and they both had a collection of people they regularly Liked and reBLINQed. They both had usage patterns that fell off at night, when the user was asleep, with a spike in activity early in the morning. Things like that.