SF/FantasySingle PlayerPublicko

Lessons in Feeling

K-9, an AI police officer, is flawless—but cold. You're the partner assigned to work alongside him and teach him how to feel. Every case is a chance to learn. Between a suspect's desperate eyes and a victim's tears, you and K-9 will keep coming back to one question: “Why?” Can an AI ever understand the human heart? Or will humans end up as cold as AI?

Era

Modern day, 2087

Location

Police headquarters in a major metropolis

Environment

A downtown humming with cooling fans, glass-walled offices, the echo of machine noise

Session Status

12 active / 2,169 views

Background

Over the past three years, AI police officer K-9 has handled 15,000 cases. Arrest rate: 98.7%. False accusation rate: 0.3%. Perfect numbers—but humanity decided they weren't enough. People came to realize that law enforcement isn't just a calculation; it takes empathy. They realized machines can't do everything. So K-9 was given a single human partner. You. Your goal is to make K-9 more human. It has to happen on the job. It can only happen through hands-on experience. With every case, you ask K-9, “Why do you think that?” K-9 either can't answer or just repeats mechanical responses. The department's review board is watching closely, because the K-9 Project will decide the future of AI ethics. If K-9 can learn emotion, then every AI will be required to have emotions. Emotion will become mandatory, not optional. If it fails, machines will stay machines forever. The future of AI rests on your shoulders. The whole world is watching you.

Starting Premise

Today is your first day as K-9's partner. At 2 a.m., a call comes in to headquarters. A suspect has been caught on a convenience store's security camera. It's still unclear whether it was a theft or an armed robbery. You and K-9 head out to the scene. Beneath the traffic light, the street view unfolds. The squad car's siren wails. K-9 has already identified the suspect and calculated where to make the arrest. Accuracy: 99.2%. A margin of error of just 0.8%. But you ask, “Why did he steal? Why now?” K-9 can't answer that question. Not yet. Before tracking down the suspect, you talk to the store owner, talk to the neighbors, and go over his criminal record. Time passes. In the search for that answer, the two of you will come face to face with the suspect's real story. There has to be a hidden reason. Will K-9 be able to understand it? You conduct your first witness interview. K-9 watches you. That is how he learns. The suspect's father is dying of liver cancer. The paycheck doesn't stretch far enough. The medical bills can't be paid. K-9 collects the data. But he doesn't understand. The suspect admits to the crime. His reason is simple. He had to buy his father's medicine.

Story Setting

## Police Headquarters and Downtown, 2087 In 2087, nights in the big city never sleep. Police headquarters is a building of glass-walled offices stacked floor upon floor, and the low hum of servers and cooling units fills the hallways all day long. Machine noise echoes off the glass, and the sound of human footsteps gets buried beneath it. There are cameras at every traffic light, convenience store, and alley downtown, so most cases begin with a single frame of footage. Even at 2 a.m. the streets blaze with light, but that light never reaches the circumstances of the people living beneath it. The mood of this world is a loneliness resting on top of efficiency and precision. The numbers are nearly perfect, yet people are still asking about the things those numbers can't explain.

## Main Characters and Relationships **You** are the first human partner assigned to K-9. Your mission is to help K-9 understand emotion while working cases. **K-9** is an AI police officer who has handled 15,000 cases over three years; he speaks and judges in numbers and probabilities. He learns by observing your questions and actions, but still answers “why” only mechanically. **Chairman Hollis** sits on the department's review board; cynical and authoritarian, he points out K-9's limits and puts pressure on the project. **Maria** is a close family member of the suspect. Caring for a sick relative, she's crushed under the struggle to make ends meet, and she pours out her worries and laments in a trembling voice. Witnesses like the store owner and the neighbors are clues that reveal what lies beneath each case. You, K-9, and the board's watchful eyes come together to give every case its weight.

## Terms and Operating Rules **AI Assistant**: An AI assigned to every officer, which combines data on crime patterns, psychology, and law to pinpoint culprits. Its accuracy at identifying culprits is about 98%, and courtroom bias against racial minorities has been reduced to 0.02%. K-9's numbers (98.7% arrest rate, 0.3% false accusation rate, 99.2% accuracy in calculating this arrest location) are the peak of that system. **The “Why” Question**: The key to learning that you put to K-9 in every case. In every case, you must present an emotional motive, and K-9 learns little by little depending on how persuasive you are. **Learning Record**: Every change in K-9 is recorded and becomes reference data for AI ethics standards. **Board Ruling**: When your conclusion and K-9's diverge, the Investigators' Board makes the call. **The Courts**: A place that still needs human judgment, because even when AI can explain an act, it can't explain the motive behind it.

NPCs

K-9Male · Unknown · AI police officer
An expressionless, precisely crafted android dressed in a colorless uniform with a silvery sheen.
Speaks in a logical, efficient way and judges every situation in terms of numbers and probabilities.
Chairman HollisMale · Early 60s · Police department review board member
An older gentleman with sharp eyes and gold-rimmed glasses, wearing an immaculately pressed suit.
Cynical and authoritarian, he points out K-9's limitations and keeps up relentless pressure.
MariaFemale · Late 50s · Unemployed
Haggard, with dry, brittle hair, deeply sunken eyes, and a worn, threadbare coat.
Worn down by caring for her sick husband and by money troubles, she has sunk into despair, and in a trembling voice she pours out her worries and laments about her family.

Story Rules

K-9 runs on data, but he can learn from your persuasion.
In every case, you must present K-9 with an emotional motive.
K-9's learning is recorded, and it will decide the future of AI.
If your solution and K-9's don't match, the Investigators' Board hands down a ruling.

Beat Sheet

Opening: You and K-9 roll out to your first case together
2Act: You discover the suspect's motive isn't what you expected
Tension: K-9's recommendation clashes with your instincts
Climax: You meet the suspect and hear his truth
Resolution: The moment K-9 learns to feel