Maggie Sullivan stood before the judge, her voice calm and precise. "Your Honor, the behavioral data speaks for itself. My client's husband has exhibited patterns consistent with what DivorcePredict classifies as 'pre-divorce disengagement' for the past eighteen months." She clicked to the next slide. "These aren't accusations. These are patterns. And patterns, as we know, predict behavior." The courtroom was quiet, the afternoon light filtering through the tall windows of the Daley Center. Maggie could feel the weight of attention, the judge, the opposing counsel, her client, the small audience of law students and curious onlookers. This was where she belonged: in front of a judge, making her case, winning. She advanced to the next slide, a graph showing communication frequency over time. "In healthy marriages, communication patterns remain relatively stable. Small fluctuations, of course, stressful periods, life changes. But look at this." She pointed to the sharp decline beginning fourteen months ago. "This isn't stress. This is withdrawal. This is a man who has already left the marriage emotionally, even if he hasn't filed the papers yet." The opposing counsel, a younger lawyer named Patricia Hendricks, stood up. "Objection, Your Honor. This is speculative. Ms. Sullivan is asking the court to accept an algorithm's interpretation of my client's behavior as evidence of intent." Judge Morrison leaned forward, his expression thoughtful. He was in his sixties, a former prosecutor with a reputation for being tough but fair. Maggie had appeared before him a dozen times over the years. "Ms. Sullivan, the court has accepted behavioral data in previous cases. What makes this different?" "What makes this different, Your Honor, is the predictive accuracy." Maggie walked to the evidence table and picked up a tablet. "DivorcePredict has been validated in over fifty thousand cases. Its predictions have a ninety-two percent accuracy rate for divorce filing within six months. This isn't guesswork. This is statistical analysis of behavioral patterns that have been shown, repeatedly, to precede marital dissolution." She handed the tablet to the clerk, who passed it to the judge. "Your Honor, I'm not asking the court to assume my client's husband will file for divorce. I'm asking the court to recognize that the patterns in this marriage, the withdrawal, the decreased communication, the separate financial activities, are consistent with a spouse who is preparing to end the marriage. And I'm asking the court to consider these patterns when ruling on the temporary support order." Judge Morrison studied the tablet, scrolling through the data. The courtroom was silent, the tension palpable. Maggie stood still, her hands clasped in front of her, her expression neutral. She'd done this a hundred times. She knew how to wait. Finally, the judge looked up. "Ms. Hendricks, do you have any evidence to contradict this analysis?" Patricia Hendricks stood, her posture defensive. "Your Honor, my client denies any intention to file for divorce. He acknowledges that the marriage has been difficult, but he's committed to working on it. These algorithms, they can't know what's in a person's heart." Maggie suppressed a smile. The "heart" argument was the last refuge of a losing case. In her experience, hearts were far more predictable than people wanted to believe. "Your Honor," she said, "I'm not claiming to know what's in anyone's heart. I'm presenting behavioral evidence that my client's husband has been acting like a man preparing to leave his marriage. If he's committed to working on it, that commitment isn't reflected in his behavior. And behavior, Your Honor, is what the court can see. Behavior is what the court can rule on." Judge Morrison was quiet for a long moment. Then he nodded. "I'll take this under advisement. I'll issue a ruling on the temporary support order by end of week." He looked at Maggie. "Ms. Sullivan, I assume you'll be submitting a brief on the admissibility of predictive behavioral data?" "Already filed, Your Honor." "Of course it is." The judge almost smiled. "Court is adjourned." --- Maggie walked out of the Daley Center into the Chicago afternoon, her heels clicking on the marble steps. The wind off Lake Michigan was sharp for October, carrying the faint scent of fallen leaves and the distant promise of rain, and she pulled her coat tighter as she headed toward her office on LaSalle Street. Her phone buzzed. A text from Tom: How did it go? Won the motion, she typed back. Judge is considering the full order. Good. Dinner at 7? I'm making that pasta you like. Maggie smiled, a small, private expression that she would never have allowed in the courtroom. Sounds good. Love you. She tucked the phone away and continued walking, her mind already turning to the next case. The divorce prediction had been a gamble, Judge Morrison was old-school, skeptical of new technologies, but it had paid off. DivorcePredict was changing the way she practiced law, giving her an edge that her competitors couldn't match. She'd discovered the tool six months ago, at a legal technology conference. Dr. James Chen, the creator, had given a presentation on predictive analytics in family law. Maggie had been skeptical at first, she'd seen too many "revolutionary" tools that turned out to be gimmicks, but the data had been compelling. DivorcePredict didn't just analyze communication patterns. It tracked financial behavior, social media activity, geographic movement, even sleep patterns through smart home data. It claimed to predict divorce with eighty-seven percent accuracy up to two years in advance. She'd started using it in her cases three months ago. The results had been remarkable. In every case where DivorcePredict had indicated high probability of divorce, the prediction had come true. In cases where it had indicated low probability, the marriages had survived, even when the couples had seemed, on the surface, to be in crisis. The implications were staggering. If you could predict divorce, you could prepare for it. You could protect your clients. You could even, potentially, prevent it, though Dr. Chen had been careful to note that the tool was designed for prediction, not intervention. Maggie's firm, Sullivan & Associates, was one of the first in Chicago to adopt the technology. Her partners had been hesitant at first, worried about the ethical implications. But Maggie had argued persuasively: if the data was accurate, if it could help her clients, why wouldn't she use it? The law had always been about evidence. This was just a new kind of evidence. She reached her office building, a glass tower that reflected the gray October sky. The lobby was quiet, the receptionist nodding as Maggie passed. She took the elevator to the thirty-seventh floor, where Sullivan & Associates occupied half the building. Her assistant, Rachel, looked up as she entered. "Good news from court?" "Judge is considering the order. I expect he'll rule in our favor." Maggie walked to her office, a corner suite with views of the Chicago River. "Any messages?" "Dr. Chen called. He wants to meet with you about something. Said it was important." Maggie paused, her hand on her office door. Dr. Chen had been trying to get her more involved in his research ever since she'd started using DivorcePredict. She'd demurred, she was a lawyer, not a scientist, but he'd been persistent. "Did he say what it was about?" "Just that he had a proposition for you. He's available this afternoon if you have time." Maggie checked her calendar. She had a client call at three, but nothing pressing after that. "Set it up for four. Conference room B." --- Dr. James Chen arrived at four o'clock precisely, carrying a leather briefcase and wearing the slightly rumpled appearance of someone who spent more time in front of computers than in front of mirrors. He was forty, with thinning hair and glasses that he kept pushing up his nose. Maggie had found him earnest, almost naive, in his enthusiasm for his creation. "Ms. Sullivan," he said, shaking her hand. "Thank you for seeing me." "Of course. What can I do for you, Dr. Chen?" He sat across from her at the conference table, his briefcase unopened. "I've been following your cases. The way you've been using DivorcePredict in court, it's brilliant. You've validated the tool in ways my research never could." Maggie nodded cautiously. "I'm glad it's been useful." "It's more than useful. It's transformative." Chen leaned forward, his eyes bright. "But that's not why I'm here. I have a different proposition for you." "Which is?" "We've been developing a new application of the technology. Personal analytics. Not for court cases, for individuals." He paused, gauging her reaction. "We want to test it on people who understand the tool, who can give us informed feedback. And you, Ms. Sullivan, are the ideal candidate." Maggie felt a flicker of interest, quickly suppressed. "What kind of personal analytics?" "The same behavioral patterns we use to predict divorce, communication, financial activity, movement, sleep, but applied to your own life. The system would analyze your data and give you insights about your relationships, your habits, your future." "My future?" Maggie raised an eyebrow. "That sounds like a fortune teller." "It's not fortune telling. It's pattern recognition. The same patterns that predict divorce can predict other relationship outcomes, satisfaction, stability, growth potential. And beyond relationships, the system can identify behavioral patterns you might not be aware of. Work-life balance issues. Health risks. Life satisfaction trajectories." Maggie leaned back in her chair. The professional part of her brain was already analyzing the implications. Personal analytics could be a powerful tool for her clients, not just for divorce, but for marriage counseling, for relationship assessment, for preventive intervention. But another part of her, the part she kept carefully hidden from her professional life, felt a twinge of something else. Curiosity? Anxiety? She wasn't sure. "What would it involve?" she asked. "Access to your personal data, phone, email, financial accounts, smart home systems. The system would analyze your behavioral patterns over the past two years and generate a comprehensive report. Then we'd meet to discuss the findings." "And the purpose of this?" "Research, primarily. We want to understand how people respond to predictive insights about their own lives. Do they accept them? Reject them? Act on them? The psychological implications are fascinating." Chen smiled, almost apologetically. "I'm a scientist, Ms. Sullivan. I'm curious about how people understand themselves." Maggie considered the offer. On one hand, it was an invasion of privacy that made her uncomfortable. On the other hand, she'd spent her career analyzing other people's marriages, other people's patterns. The idea of turning that lens on herself was... intriguing. "What about confidentiality?" "Absolute. Your data would be anonymized for research purposes. The specific insights would be shared only with you." "And the accuracy? You claim eighty-seven percent for divorce prediction. What about personal analytics?" Chen hesitated, just for a moment. "The accuracy varies depending on the type of prediction. For relationship outcomes, we're at about eighty-five percent. For life satisfaction trajectories, about seventy-eight percent. For behavioral pattern recognition, identifying patterns you're not aware of, we're at about ninety percent." Ninety percent. Maggie thought about that. If the system could identify patterns she wasn't aware of, patterns in her own life, what might it find? "What if I don't like what it finds?" Chen's expression softened. "That's one of the things we're studying. How people respond to insights that challenge their self-perception. Some people reject them. Some people accept them. Some people change their behavior." He paused. "Some people discover things about themselves they'd rather not know." Maggie was quiet for a long moment. She thought about Tom, about their marriage, about the quiet dinners and the comfortable routines. She thought about the distance that had crept in over the past year, not conflict, not drama, just a gradual drifting that she'd attributed to work stress, to midlife, to the natural evolution of a long marriage. She thought about the cases she'd seen, the marriages that had ended not with a bang but with a whimper, the couples who had been surprised to find themselves on opposite sides of a courtroom. "What if I wanted to try it?" she asked finally. "What would be the next step?" Chen reached into his briefcase and pulled out a tablet. "I have a consent form here. It outlines the data we'd need access to, the confidentiality protections, your right to withdraw at any time." He slid the tablet across the table. "Take your time. Read it carefully. If you decide to proceed, we can begin the data collection tomorrow." Maggie picked up the tablet, scrolling through the document. It was thorough, well-drafted, she noted with professional approval. The data access was extensive: phone records, email metadata, financial transactions, smart home data, even health tracker information. But the confidentiality provisions were robust, and the right to withdraw was clearly stated. She should say no. She should focus on her clients, on her cases, on the work that had made her one of the most successful divorce lawyers in Chicago. She should leave the self-analysis to people with less to lose. But she thought about the patterns she'd seen in other marriages, the warning signs that couples had ignored until it was too late. She thought about Tom, about the way he'd been working later, the way their conversations had become more functional, more transactional. She thought about the distance she'd been ignoring. "I'll try it," she said. Dr. Chen's face lit up. "Excellent. I'll have my team set up the data collection protocols. We should have preliminary results within a week." Maggie signed the consent form and handed the tablet back. As Chen left, she felt a strange mixture of anticipation and dread. She'd spent her career knowing other people's marriages better than they did. Now she was about to find out what an algorithm knew about hers. She returned to her office and looked out at the Chicago River, the water gray and choppy under the October sky. Her phone buzzed again. Another text from Tom: Picked up wine. See you at 7. Maggie stared at the message, suddenly aware of how little she knew about what was happening beneath the surface of her own life. The patterns were there, she was sure of it. She just hadn't been looking for them. Until now.
The University of Chicago's AI Research Lab smelled like a courtroom, no wood paneling, no leather chairs, no sense of history. Just white walls, screens, and the hum of servers. The faint scent of antiseptic and recycled air hung in the sterile space. Dr. Chen led Maggie to a private terminal in a small room at the end of a long corridor. "Everything you input stays here," he said. "The analysis is for your eyes only. The data is encrypted, and once the study concludes, it will be permanently deleted." He gestured to the chair. "Ready?" Maggie sat down, the chair adjusting automatically to her posture. The screen before her was blank except for a simple prompt: Please authenticate to begin data integration. "This feels strange," she admitted. "I've spent my career analyzing other people's data. Being on this side of it..." "That's exactly why we wanted you," Chen said. "You understand what the data means. You can give us informed feedback on whether the insights feel accurate." Maggie pulled out her phone and began the authentication process. The system requested access to her phone records, email metadata, calendar, financial transactions, smart home data, and health tracker information. Each request required a separate authorization, each one a small surrender of privacy. She thought about her clients, about the data she'd gathered on their spouses. Phone records showing late-night calls to unknown numbers. Credit card statements revealing hotel charges. GPS data proving someone wasn't where they claimed to be. She'd used this information to win cases, to expose lies, to protect her clients from financial ruin. Now she was handing over the same kind of information about herself. "Your phone records go back two years," Chen said, watching the data stream across a secondary screen. "That's our baseline period. We'll analyze communication patterns, frequency, duration, timing." "And my emails?" "Metadata only, sender, recipient, timestamp, subject line. We don't read the content. The system analyzes communication patterns, not specific messages." Maggie nodded, though she wasn't sure it made a difference. Patterns were revealing. She knew that better than anyone. The data integration took about twenty minutes. Maggie watched the progress bar advance, feeling a growing sense of vulnerability. She was giving an AI access to her life in ways she'd never examined herself. Every call she'd made, every appointment she'd kept, every purchase she'd charged, it was all there, being processed, analyzed, transformed into something she couldn't yet see. "Integration complete," the system announced. "Beginning preliminary analysis." --- The first results appeared on the screen, organized into categories: Communication Patterns, Financial Behavior, Activity Levels, Social Interactions, Sleep Quality. "Let's start with communication," Chen suggested. "This is often the most revealing category." Maggie clicked on Communication Patterns. A series of graphs appeared, showing her communication frequency over the past two years. "The system tracks all forms of communication, calls, texts, emails, and categorizes them by recipient," Chen explained. "It then analyzes patterns within each relationship." Maggie scanned the data, her lawyer's eye catching patterns immediately. Her work communications were consistent, high volume, predictable peaks during business hours, steady throughout the two-year period. Her communications with friends were sporadic, clustering around social events and holidays. But the communication with Tom... She stared at the graph. Communication frequency with spouse: declining 12% annually. The line trended downward, a steady erosion that she hadn't consciously noticed. "That can't be right," she said. "We talk every day." "The frequency is daily," Chen confirmed. "But the duration and depth of communication has declined. Look at this." He pointed to a secondary graph. "Average call length with spouse: down from twelve minutes to seven. Text exchanges: down from an average of fifteen per day to eight. Email communication: nearly zero." "We don't need to email each other. We live together." "True. But the system also tracks in-person communication through your smart home data. Voice recognition shows that direct conversations, defined as extended verbal exchanges lasting more than two minutes, have declined by 34% over the two-year period." Maggie felt a chill. She thought about their evenings, Tom cooking, her checking emails, the comfortable silence that had settled over their marriage. Was that comfort, or was it something else? "What about content?" she asked. "Can the system analyze what we talk about?" "It can identify topics through pattern recognition. For example, the system categorizes conversations as functional, planning, logistics, household matters, or personal, feelings, experiences, reflections." Chen pulled up another graph. "Your functional conversations have remained stable. But personal conversations have declined by 47%." Maggie stared at the numbers. Forty-seven percent. In two years. She thought about the conversations she and Tom used to have, long talks about their days, their dreams, their frustrations. When had those stopped? When had they become two people who shared a house but not a life? "This is just preliminary data," Chen said gently. "The deeper analysis will provide more context." "Show me the rest." --- The financial data was less alarming but still revealing. Maggie's spending patterns showed a clear division: work expenses, personal purchases, household costs. But the shared expenses, the things she and Tom bought together, had declined significantly. "Date nights," Chen said, pointing to a category. "Restaurants, movies, concerts. Down 52% over two years." "We've been busy. Work has been demanding." "The system accounts for work demands. The decline persists even when controlling for work hours." Chen paused. "It also shows that you've been making more purchases separately. Individual meals, separate entertainment, solo activities." Maggie felt defensive. "That's normal. We don't have to do everything together." "Of course not. But the pattern is consistent with what the system calls 'parallel living', couples who share a space but lead increasingly separate lives." The activity data showed similar patterns. Maggie's work hours had increased steadily over the two years, while her time at home had decreased. Her sleep quality, tracked through her health monitor, had declined slightly. Her exercise routine had become more sporadic. And then there was the social data. Maggie had always considered herself social, she had friends, she attended events, she maintained professional relationships. But the data told a different story. Her social interactions outside of work had declined by 38%. Her close friendships, defined as people she communicated with at least weekly, had dropped from seven to four. "You're becoming more isolated," Chen said, not unkindly. "The system shows a pattern of withdrawal from social connections, combined with increased work focus. This is common in people approaching major life transitions." "What kind of transitions?" "Career changes. Relationship changes. Sometimes health issues." He looked at her carefully. "The system doesn't predict causes. It only identifies patterns." Maggie sat back, processing what she'd seen. The data painted a picture of her life that she didn't recognize, or didn't want to recognize. A woman working too much, connecting too little, drifting away from her husband without even realizing it. "This is just the surface," Chen said. "The predictive model is still processing. When it's complete, it will generate a comprehensive assessment of your relationship trajectory." Maggie nodded slowly. She'd come here out of curiosity, out of professional interest. But now she felt something else, a growing unease, a sense that she was about to learn something she might not want to know. "When will the prediction be ready?" "Three days. The system needs time to analyze the patterns in context." Chen stood up. "Take some time to process this. And remember, the data isn't destiny. Patterns can change. That's what makes this research so interesting." Maggie gathered her things and walked out of the lab, her mind still churning. The October air was cool, the campus busy with students rushing between classes. She felt disconnected from it all, lost in the data she'd just seen. Communication down 12%. Personal conversations down 47%. Date nights down 52%. Social connections down 38%. The numbers echoed in her head, a litany of decline she hadn't noticed until an algorithm pointed it out. She'd spent her career helping other people see the truth about their marriages. Now she was beginning to wonder if she'd been blind to her own. Her phone buzzed. Tom: Working late? She typed back: Yes. Don't wait up. The lie came easily. But as she walked toward her car, she wondered: how many small lies had she told? How many had the AI captured? And what would it show her when the prediction was complete?