Prediction

by Tony Batton

  He picked up Eve’s laptop and typed Millie Wright into Google. Hundreds of hits came back, all of which seemed to be about other Millie Wrights. There was too much data.

  Searches for ‘Millie Wright blogger’ and ‘Millie Wright journalist’ both immediately found her blog. He’d already skimmed through it before, but what he really wanted was something she didn’t necessarily curate. Something that might show more of who she really was. But how to find it amongst too much data.

  On a whim, he typed in ‘Millie Big Data’.

  A YouTube video of someone giving a lecture popped up at the top of the list. It was a poor quality recording, likely filmed covertly. The presenter’s name was listed as Milligram. But even though she wore a cap and dark glasses, he was immediately sure it was Millie. She stood at a lectern under patchy lighting. When she spoke it sounded like the microphone had been positioned in a tin can.

  "So," Millie said, her eyes sweeping the room. "Big Data: Big Danger or Big Opportunity? Quite a title." She paused, taking a sip of water. "And I’m supposed to respond to it. Perhaps unsurprisingly for a Big Data analysis, I’m going to tell you that there’s more than one answer."

  There were a few laughs from around the room.

  Millie nodded and continued. "When I talk about Big Data, I mean seeing and understanding the relations within and among pieces of information that, until very recently, we struggled to fully grasp. I’m talking about datasets so large and complex that traditional analytical techniques are inadequate. I’m talking about a volume of data not only beyond proper human comprehension, but also beyond conventional computing."

  She put both hands on the lectern. "What’s more interesting is what Big Data means. Scientists are using algorithms and predictive analytics to see patterns in human behavior: we can apply lessons from the past to predict the future. A decade ago that might have sounded ridiculous. But the experience of progress in weather forecasting has shown us how analysis of greater volumes of data can produce powerful results. The potential benefits are wide ranging. For instance, in identifying consumer habits: companies can use Big Data to understand customer preferences, anticipate future behaviour and develop individualized marketing campaigns. Big Data has also been used to identify infections in premature infants before symptoms appear by monitoring a range of vital signs, then correlating between minor and major problems. Companies and government entities are using Big Data to improve internal operations and reduce costs; for example, a number of cities across the world are using Big Data to determine which buildings are most at risk of fire damage."

  She cast her eyes around the quiet room. "A key point that bears repeating is that Big Data is about what, not why. As human beings we’re conditioned to look for causes – but searching for causality is often difficult and may lead us down the wrong path. With Big Data we put that aside to discover patterns and correlations that offer us novel and invaluable insights. In many situations this is good enough. For example, a change in a treatment regime leading to an improvement in the health of cancer sufferers may be perfectly acceptable, without knowing how or why it worked. Before Big Data, our analysis was usually limited to testing a small number of hypotheses defined well before we even collected the data." Millie paused, as if for emphasis. "When we let the data speak for itself, we can discover connections that we never thought existed."

  "What are the challenges in making use of Big Data? Getting data isn’t the problem: we already have too much, including things we never used to think of as information at all, such as a person's location, the vibrations of an engine, or the stress on a bridge. The real challenge is analysing it in a meaningful way, because it is beyond our current standard computing capability: it will require exceptional technologies to overcome the challenges of scale and nature.

  "This leads on to my next point: what are the risks? Or to put it another way, why would we not do it? As Big Data makes increasingly accurate predictions about the world and our place in it, we may not be ready for its impact on our privacy and our sense of freedom. For example, what if we use Big Data predictions to judge and punish people even before they've committed a crime? Handled responsibly, Big Data is a useful tool of rational decision-making. Wielded unwisely, it can become an instrument of oppression.

  "With Big Data, the new value of information resides in secondary uses – most of which weren't even imagined when the data was first collected. As for privacy, it may become an illusion. Data is often alarmingly easy to de-anonymise. Surveillance has only gotten easier too, as well as cheaper and more powerful. The ability to capture personal data is often built into the tools we use every day, from websites to smartphone apps. The data-recorders that are in most cars to capture all the actions of a vehicle a few seconds prior to airbag activation have been known to 'testify' against car owners in court." Millie gave a wry smile at a few gasps from around the room.

  "And what if we lose control? Especially if decisions are made not by humans but by machines. Computer systems always used to base their decisions on rules they had been explicitly programmed to follow. Thus, when a decision went awry, we could go back and figure out why the computer made it. It used to be that all computer code could be opened and inspected, and those who knew how to interpret it could trace and comprehend the basis for its decisions, no matter how complex. With the advent of machine learning, powered by Big Data, this is changing. Traceability will become much harder. The basis of an algorithm's predictions may be too intricate for people to understand."

  Millie paused to look around the room, meeting the eyes of her audience. "Thus, we should consider whether we have a choice in all this. Like other powerful technologies and disruptive human innovations, predictive analytics is essentially amoral, and can be used for good or evil. So should we embrace it, ignore it or fight it? Do we even have a choice? I don’t think we do. The stable door is swinging free and the horse has bolted. Pandora’s Box is open. We cannot deny that people will exploit Big Data, so we may as well look for the good we can also get out of it. Prediction is power, not least in terms of killer competitive advantages.

  "Yet, as I’ve said, perfect prediction is impossible. But that won’t stop us trying. If we chose not to explore Big Data and its analysis, it is inevitable that others will do so. And those that do will develop an advantage. But", Millie gave a wry smile, "even Big Data can't predict how Big Data will evolve. Nor can it predict the intuitive leaps that humans are capable of and the impact these have on the world. If Henry Ford had queried Big Data algorithms for what his customers wanted, the result would have been 'a faster horse'. Big Data is going to revolutionise the world, but it shouldn’t become our world. Now, are there any questions?"

  A woman off-camera called out, "I heard you were involved in the development of just such a project."

  Millie shrugged. "You may wish to check the integrity of your data collection methodology."

  "You’re advocating spying on people?" shouted a man.

  "No," Millie replied. "That was definitely not my message. This isn’t spying, certainly not in any targeted sense. The data exists. We’re just processing it in a new, if profound, way."

  "So you want a Big Brother Society?" asked the same man.

  "I'm not a politician. Just an observer. Of course, some might argue that if you’re doing nothing wrong, what’s the problem with being watched?"

  A woman closer to microphone asked, "What’s your prediction for the future?"

  Millie paused and smiled. "The demise of the expert: expertise is appropriate for a small-data world where one never has enough information, or at least not the right information, and we have to rely on intuition and experience. Now I'm sorry, but we're just about out of time. I’ll end, if I may, with a piece of someone else’s wisdom: ‘Prediction is difficult, especially if it's about the future’."

  There was a round of applause and the recording ended.

  Michael stared at the laptop screen. Millie wasn’t just a
hack journalist. He went back to her blog, MillieOnTruths.com. He had glanced at it before, but this time he read it thoroughly. There was a lot of content: well-written, well-presented, and authoritative. Under a section named ‘hot topics’ were some recent entries. Millie was appealing to her community of readers for leads relating to ‘Project Darwin’. Was this to do with ZAT?

  Michael ran a quick Google search, but produced so many hits related to the naturalist Charles Darwin that he quickly added ‘+ ZAT’. It produced nothing useful. Where else could he look?

  Michael glanced at his briefcase. Inside was his Infinity laptop.

  His fingers trembling, he slipped it from the bag and booted it up. Placing the headset on, he clicked through security then pulled up Scope, the bespoke search engine. His fingers flew over the keyboard. On his ears the headset hummed quietly.

  And there was the result. The link he had suspected: Project Darwin. Attribution: ZAT Systems Limited. Hardware/software project. Project discontinued fifteen years ago.

  Michael sucked in his lower lip. Why would Millie be particularly interested in a project terminated fifteen years ago? Below the initial report some further search responses were slowly appearing. Michael knew that a search like this, without the might of Google’s servers behind it, could take considerable time. He scrolled down. And froze. A name had appeared.

  Craig Adams. Owner: Project Darwin.

  Had his father worked at ZAT? His mother had never mentioned it, but then she never talked about his father’s work at all. Perhaps it was another Craig Adams?

  Michael took a sip of wine, and thought through his options. He could hire a private investigator. That would be expensive and slow. He also doubted they would find much more than he had; there was hardly going to be a ready paper trail. Or he could ask his mother. It was not an option that appealed. If only his father had left behind some records… but his mother had literally just cleared everything out. And yet what if Craig had hidden it? Perhaps on a computer.

  Michael sat up, almost spilling his wine, and swore loudly. The laptop. The damn laptop that his mother had just sold off. If there was a secret to be uncovered, it might have been on the device. Was there any way to get it back?

  His mother had sold it online. Perhaps he could trace the sale, then offer the buyer more than what they paid to sell it back. It was worth a shot. He logged onto eBay and ran a search for his mother's account; he had helped her set it up a few years back and soon found she hadn't changed the password. He flicked his eyes over the short list of her recent sales to the laptop. Then he looked at the buyer's details. And froze again.

  MillieOnTru.

  Millie had bought his father's laptop? Had she known what it was?

  One way or another he was going to find out. He thought about calling her, but it would be too easy for her to simply put the phone down. But he knew where she lived from the file the firm had provided.

  He pulled on his coat and was out the door in less than a minute, heading towards the nearby taxi rank.

  Fifty-Three

  At ZAT’s regional facility outside Reading, there was a discrete but increased level of activity. Gregory Jenson stood watching a facilities team bringing a large number of grey unmarked boxes and silver flight-cases into the temporary lab, placing them neatly to one side.

  Chow was crouched, checking each box and case as it was deposited. "That’s everything," he said, standing up, his knees cracking audibly. "Although I am still waiting for Kelly to get back to me on a couple of points."

  Jenson walked over to one of three long white lab tables and ran his fingertips across the top. "Astrid’s offsite for a day or two on a special project. I’ll get someone else to handle it for you. I assume you’re happy with the new set up, and not just because your commute is better?"

  Chow scratched his bald head. "I hadn’t meant that to reach your ears."

  "Best not to say anything these days that you don’t mind being passed to an unintended audience: someone’s always listening." Jenson patted him on the shoulder.

  Chow took a deep breath. "I’m delighted to be working out of this location, but I heard a rumour that the project has been cancelled."

  "Where did you hear that?"

  "It doesn’t matter. As you say, someone’s always listening. So is it true?"

  Jenson hesitated. "There are some differences in thinking between Saxton and myself, but we’re working on resolving them. You just keep doing what you do."

  "Good." Chow said. "Because I have an update for you on the software development." He opened up a nearby laptop and set it on the nearest table. "Look at this simulation of the system in operation." He pressed a key and a program started to run.

  Jenson watched the display and his eyes widened. "How is that possible?"

  Chow coughed. "It wasn’t even a software issue. I checked the circuit diagrams for the core processor."

  "Which ones?" Jenson asked.

  "All of them."

  "There are billions of circuits. How could you possibly do it in such a short period of time?"

  "I looked for patterns. And I found a logic error. A feedback loop. It was replicated in a number of places; when the blueprints were produced, someone must have pasted it in by mistake."

  "So what's our next step? Do we have to get the chips re-pressed?"

  "I can program around this. It will only be about 95% efficient, but it will work."

  Jenson nodded. "Get the system built then, when we get the decision on this project overturned, we'll be ready to move."

  In a grey van parked across the street from the ZAT Reading building, two men sat apparently reading the newspaper and eating sandwiches. The taller man flickered his gaze at the entrance, watching Gregory Jenson make his way out to his parked car. The shorter man was engrossed in a laptop running a detailed analysis routine, parsing data from a series of antennas that had been discretely deployed around the building. He peered more closely at the display, grunted and tapped the headset hooked over his ear.

  "Ms Rose, progress is good. We almost have a full sweep. After that we just need to pull the team together. We should be ready in forty-eight hours."

  Fifty-Four

  Michael knocked loudly on the painted door of the apartment. There were footsteps inside, then a mutter of surprise. The door opened a crack, held in place by a chain.

  "What the crap?" Millie asked, just visible through the gap. "Are you stalking me?"

  Michael took a step back. "We need to talk."

  "Then what’s wrong with the café like last time? How do you even know where I live?"

  Across the hall another door opened and a large man stepped out, glaring at Michael. "What's going on?"

  Millie sighed and unchained the door. "It's OK. He's a… friend. Obviously had too much to drink." She grabbed Michael's arm and pulled him into her apartment. Closing the door, she walked into a small lounge and pointed at a sofa. "Have you been drinking? What couldn’t wait?"

  Michael took the seat. "Do you know who Craig Adams is?"

  She frowned. "I do. Why?"

  "Did you not go away and do any research on me?"

  "I might have got round to it. You didn’t seem that important. No offence." Millie stared at him, then realization appeared on her face. "Adams. He was your father?" She glanced around. "Look, it’s not that unusual a surname."

  "I suppose not. What can you tell me about him?"

  "You want me to tell you about your father? Shouldn’t it be the other way around?" She hesitated. "If this is some kind of trick—"

  "I’m not asking for my firm. I’m asking for me."

  She glared at him. "Are you recording this conversation? Because I know you have to ask permission."

  "I’m not recording anything. Look, I'm sorry for how I approached you before, but I was just doing my job."

  "As was I."

  "And I would very much like to know what you learned while doing it. About my father." Michael
took a deep breath. "I read your blog. You mentioned the Darwin Project."

  "How did you connect that name to your father?"

  "I don’t know anything about his work. But I know you have my father's laptop. You bought it off eBay a week ago."

  "That was your father's?"

  "I'll pay you twice what you paid for it."

  "I didn’t… I don't care if you offer me a hundred times what I paid. You can't have it." She raised a hand. "But you can look at it with me if you like. And, realistically, I may be one of the few people who can actually get it to work."

  "I think I know how to use a computer."

  "This is a different kind of beast. Without my help, it’s a lump of plastic and wire. But I’m starting to think it may contain something important because your father was, if my research is accurate, an important man."

  "My mother said he was a dreamer." Michael rubbed his temples. "This is just a game for you. But this is my father."

  Millie gripped his hand. "Then let me help you get to the truth."

  Fifty-Five

  Michael and Millie sat in the tiny spare bedroom that served as her office, staring at his father’s laptop.

  "Bespoke design, built from non-standard components that are at least fifteen years old." Millie ran her fingers lightly over the casing. "I haven't switched it on yet, so you haven't missed anything."

  "I could never get it to boot up."

  "Because you're not an IT genius." she said, sitting up straight. "The power supply and the battery are dead. Nothing you can do about them, so I've built a replacement. The voltages are non-standard, just to add some spice." Millie picked up an odd-looking power adaptor and, with a flourish, plugged it in.