Rulescuck
Prediction Markets
Marcus watched Sarah squeeze the toothpaste from the middle of the tube, and his heart sank.
“Come on,” he muttered, refreshing his portfolio. The market had already resolved: Will Sarah Chen squeeze toothpaste from the bottom of the tube? NO. Confidence: 99.7%.
He’d bet yes. She always squeezed from the bottom. Three years of archived streams had confirmed the pattern. But this morning, inexplicably, she’d grabbed the tube differently, her thumb pressing into the center, and the Oracle, the adjudication AI that watched her every move through seventeen billion concurrent feeds, had ruled against him in 0.003 seconds.
Minus forty credits.
On his wallscreen, Sarah rinsed her brush and smiled at the camera drone hovering near her bathroom mirror. Twelve billion people were watching her brush her teeth at 7:47 AM Pacific. Marcus was one of perhaps three hundred million who’d placed bets on it.
“Good morning, everyone,” she said, her voice carrying the practiced warmth of someone who’d been famous since she was sixteen. She was twenty-nine now, the most-watched human being in history, her entire life converted into a continuous stream of prediction opportunities. “I’m feeling spontaneous today. Who else is feeling spontaneous?”
Marcus felt his stomach turn. Spontaneous was bad. Spontaneous meant variance. Spontaneous meant the models couldn’t track her.
But Marcus didn’t use models. He used knowledge. He’d memorized her sleep schedule, her dietary preferences, her relationships with her producers and managers and the revolving door of romantic interests that tabloids tracked with algorithmic precision. He knew her better than anyone, he told himself. Better than her own mother, probably.
The breakfast market opened at 7:52 AM. What will Sarah Chen eat for breakfast? Options: Avocado toast (34%), Smoothie bowl (28%), Eggs and bacon (18%), Other (20%).
Marcus slammed two hundred credits on avocado toast. It was Tuesday. She always ate avocado toast on Tuesdays. The correlation was perfect across eighteen months of data he’d personally compiled.
Sarah wandered into her kitchen, camera drones repositioning to catch every angle. Her personal chef, a quiet man named Eduardo who never appeared on stream but whose presence was inferred by the quality of her meals, had left several options on the counter.
She picked up a banana.
Just a banana. She peeled it, took a bite, and walked out of the kitchen.
Market resolved: Other. Banana. Confidence: 99.9%.
Marcus stared at his screen. Minus two hundred credits.
“A banana?” he said aloud, his voice cracking. “She’s never just eaten a banana. Not once. Not in three years.”
But the Oracle didn’t care about historical patterns. The Oracle only cared about what it observed, processed through neural architectures so vast that their decision-making was effectively instantaneous and, more importantly, unchallengeable. When the Oracle resolved a market, the resolution was final. No appeals. No human judgment to corrupt the outcome. That was the whole point, the innovation that had made prediction markets viable at scale. You couldn’t argue with an AI that processed visual data at the speed of light.
By noon, Marcus was down eighteen hundred credits. Sarah had taken an unexpected phone call (he’d bet against interruptions), worn a blue dress instead of her predicted green, and expressed mild annoyance at her manager, a reaction Marcus had correctly anticipated but bet on the wrong market, accidentally placing his wager on “will she be annoyed by a studio executive” rather than “will she be annoyed by her manager.”
The Oracle had ruled the manager was not a studio executive. Technically correct. Financially devastating.
He kept betting. He couldn’t stop. The next market would be different. The next market, his knowledge would pay off.
By six PM, he was down nine thousand credits. His savings.
By eight PM, the number was twelve thousand. He’d dipped into his emergency fund, then his retirement account, then a high-interest credit line he’d never intended to use.
Sarah was winding down her evening stream, doing her skincare routine, the viewership swelling as the most popular market of the day approached: When will Sarah Chen fall asleep? Over/under: 11:15 PM.
The betting volume spiked. Four hundred billion credits changed hands in the first thirty seconds of the market opening. Marcus watched the odds shift, watched the aggregated wisdom of twelve billion degenerate gamblers converge on a consensus.
He had eleven thousand credits of borrowed money.
He put it all on 11:47 PM. She always fell asleep at 11:47. Always. It was the one constant in her chaotic, livestreamed existence.
Sarah yawned, earlier than usual.
“I’m so tired today,” she murmured to her audien
ce, climbing into bed at 10:58 PM. “I think I might actually get some good sleep tonight.”
Marcus stopped breathing.
She couldn’t. She wouldn’t. Not tonight. Not when everything depended on it.
He was in his car before he realized he’d made a decision. Her address wasn’t public, but three years of obsessive viewing had given him enough contextual clues to triangulate. The Oracle tracked her location in real-time for the transportation markets; he’d learned to read between the data points.
Twenty-three minutes later, he was standing at her bedroom door, having bypassed security systems that were designed to stop stalkers with weapons, not desperate men with nothing left to lose.
Sarah sat up in bed, eyes wide and terrified, camera drones swiveling toward the intrusion.
“Don’t,” Marcus demanded. “Please. Just stay awake until 11:47. That’s all I’m asking.”
She stared at him, fear in her eyes. Twelve billion people stared at him on stream.
The Oracle opened a new market: Will the intruder be arrested and/or killed within the next thirty seconds?
Betting volume spiked.