In the witness stand, Professor Chen snapped his fingers, and a three-dimensional projection of a city appeared in the courtroom’s center, floating like a ghost town. Tiny drones buzzed like gnats, cars crawled like ants, and people drifted like specks of dust. Strangely, the skyscrapers, bundled like clusters of chopsticks, all emitted vibrant fireworks from their rooftops. It looked more like a bundle of dynamite with lit fuses than a celebratory scene.

Professor Chen spread his hands, and the image zoomed in on an intersection marked Cecil Street and McCallum Street. Pedestrians stared upward, pointing at what might be a rooftop party. Many pulled out their phones to take pictures or record videos, smiles of surprise lighting their faces as if receiving an unexpected gift.

The scene fast-forwarded: a shadow fell from the sky, landing heavily in front of a convenience store at the intersection. The crowd gasped and scattered.

The fallen figure was a person, backpack strapped on, waist bulging with equipment, blood pooling beneath them.

Among the onlookers, the face of one man was framed in red. Professor Chen gestured, freezing the frame. He spread his hands apart, and the man’s head enlarged rapidly, dominating the courtroom. It was the same head as the defendant sitting in the dock—Luo Jia-bao. Compared to the projection, his actual head seemed shrunken and diminished.

Professor Chen, the prosecution’s expert witness, began to speak:
“The ‘HeartNet’ emotion analysis technology operates by measuring the contraction and relaxation of fifty facial muscles. Each emotion activates different muscle groups with varying intensity. For example, when someone laughs wholeheartedly, it involves these muscle groups…”

Luo Jia-bao’s projected face displayed colored lines and numbers, eerily twisting into a smile.

“…whereas crying in sorrow involves this set.” The head began to weep.

“In other words, facial expressions can be quantified. By converting expressions into data and inputting it into a neural network, we receive outputs categorizing the type and intensity of emotion. While basic emotions like laughter and tears are discernible to the naked eye, extreme emotions—such as profound joy versus deep sorrow, or shock versus rage—can appear strikingly similar. Without context, distinguishing between them becomes subjective. That’s where AI ensures objectivity.”

“Professor Chen,” the prosecutor asked, “please tell us the system’s findings. When the person fell and was injured, what emotions did Luo Jia-bao’s face exhibit?”

“Concern, anxiety, and sorrow.”

“And what about the other onlookers?”

“They showed fear and anxiety, but no concern or sorrow.”

“How do you explain Luo Jia-bao’s unusual emotional response?”

“We cross-referenced the data with similar cases. Luo Jia-bao’s emotional pattern matches closely with that of a family member witnessing a critically ill loved one. Thus, two possibilities exist: either the defendant is related to the fallen person, or he empathizes deeply with them.”

The prosecutor challenged him. “Professor, isn’t your conclusion rather presumptive? Must concern always indicate a personal connection? Can’t someone simply show compassion for a stranger out of basic humanity?”

“I speak only through data,” Professor Chen replied. “Over the past twenty years, we’ve analyzed footage from countless accidents. Our findings consistently show that people react to injured strangers with fear or hesitation, never detectable concern—unless Luo Jia-bao is a unique outlier among millions.”

The projected head returned to normal size, and the street scene resumed. The injured person struggled to stand and revealed their face: a masked protester.

A police officer entered the frame, drawing his gun and shouting.

The following sequence played in slow motion. Upon seeing the officer draw his weapon, Luo Jia-bao rushed forward, hands raised, placing himself between the officer and the fallen protester. Simultaneously, the protester reached into his waistband. The officer, seeing this, adjusted his stance, finger tightening on the trigger.

In that tense, critical moment, Luo Jia-bao lunged, grappling for the officer’s gun.

---

Lin An-yi!

The salesperson called her name twice before Lin An-yi snapped out of the news broadcast. The scene of Luo Jia-bao seizing the gun still lingered in her mind like a ghost, flitting as her gaze shifted.

She had seen this footage multiple times across various media outlets. The first time was through an official photo release. Luo Jia-bao, a frail middle-aged man whose skinny frame resembled that of a malnourished teenager, stood in stark contrast to the armed officers, whose bodies were rigid as stone. To attack law enforcement with such a feeble physique—was it recklessness or a calculated risk, betting on the officers’ restraint?

Later, she saw footage recorded by bystanders. Luo Jia-bao moved diagonally, positioning himself between the officers and the fallen person. His hands were raised, as if saying, “No, stop” or “Wait, wait.” The moment an officer seemed ready to fire, he swiped at the gun with his right hand. The officer dodged, and startled, Luo grabbed the weapon with both hands. Netizens analyzed the footage frame by frame, concluding that Luo’s intent was likely to deflect the weapon to protect the fallen individual, but in the heat of the moment, it escalated into a struggle for the gun. Besides, police firearms were fingerprint-locked—useless in another’s hands.

Then the narrative shifted again. Pro-government media magnified Luo’s every move, overlaying color codes, arrows, and explanatory text to cast his actions as aggressive. His raised hands were labeled as a prelude to seizing the weapon. He was doxxed: unemployed, socially discredited, socially awkward, with a history of domestic violence, and known for his anti-police sentiments.

This became the crux of the legal arguments: guilt was inevitable, but the charges and sentencing hinged on his motives. Assaulting an officer and obstructing justice were indisputable, but if he could prove no intent to rob, perhaps he could avoid the harsher charge. The most severe threat, however, was the prosecution’s attempt to frame him as a collaborator in subversion, invoking national security laws.

Lin An-yi typically found news of this kind uninteresting. No, she was numb to all news. Her mind and emotions were encased in a hard shell—she neither ventured out nor allowed anything in. Yet the day of Luo’s incident coincided with the “Rooftop Fireworks Festival.” Citizens spontaneously flooded the streets, singing and dancing in wild jubilation. Where was she that day? At school. All the teachers and students had rushed to the corridors, the playground, and the parking lot to point at the buildings, which erupted like giant sparklers. Even she, who had been lying lazily on her desk, got up and stood on tiptoe behind her classmates to watch.

“This is your Conch smart earbud,” the salesperson said, lifting her hair and gently inserting the spiral-shaped device into her ear canal. “I’ll help you test and fine-tune it.”

That day’s atmosphere had been manic. From afar, the sounds of singing and dancing grew louder, like an advancing flame. She saw older students gather on the basketball court, unbuttoning their shirts and breakdancing. Teachers were on their phones or scrolling the internet; some lit cigarettes, others flirted and joked. The usual strict discipline, as rigid as the school’s walls, crumbled into chaos. It was as if everyone tacitly agreed to prolong this rare moment of disorder for as long as possible.

“All set,” the salesperson smiled. “The Conch has successfully measured your brainwaves, pulse, and temperature. You’re slightly excited, but don’t worry—that’s common for first-time users.”

That day, all taboos were lifted. She saw two classmates kissing behind the classroom door. Some students climbed over fences and disappeared into the streets. The parking lot echoed with alarms—someone had smashed a teacher’s car window. Suddenly, colored smoke erupted from the administrative building, and someone darted across the rooftop. She heard footsteps pounding overhead, followed by the whir of drones. There was the crack of an electric stun round, and the air filled with the acrid scent of ozone. Her attention followed the noise to the end of the corridor, where she saw bullies dragging a girl by her hair toward the restroom.

Something inside her broke free. She moved toward the restroom, step by step. What happened next was stored in her memory like crime scene photos: the flowerpot on the walkway, the bully’s head, the shattered pot, the blood splattered everywhere, and the terrified faces.

“See this brainwave pattern?” The salesperson continued cheerfully, “Smooth and undulating—it’s a signal of comfort and joy. Thank you for using our service.”

Her scattered gaze refocused. The salesperson smiled warmly. “Now let me explain the features and usage.”

“HeartNet is a typical artificial neural network designed to classify and interpret emotions. It’s also a subsystem of SkyNet, with direct access to its massive datasets. By linking facial expressions captured by SkyNet’s cameras nationwide with personal social media activity and contextual tags, it has trained its neural network through billions of reverse-propagation cycles to achieve minimal error.” Professor Chen always grew animated when discussing technical details.

“Why is the Conch so precise at detecting emotions? Because it doesn’t just rely on physiological signals—it also incorporates contextual analysis. It factors in the user’s activity, location, and companions. Most importantly, its AI is directly connected to HeartNet, which aggregates the collective wisdom of the entire population. So, it sometimes understands you better than you understand yourself. In other words, it knows you better than you do.” The salesperson’s enthusiasm grew. “We often think we understand ourselves best, but humans are notoriously unclear about their own emotions, aren’t they?”

Lin An-yi nodded. She found the statement profoundly accurate—a perfect diagnosis of her condition.