Ethical Simulator: Dilemmas of the Future

You are the AI Ethics Team. Read the following dilemmas and decide how you would program/desing the AI system.

You must justify your decision with one ethical principle.

________________________________________

Dilemma 1: The Autonomous Drone

Your autonomous drone has discovered two survivors of a crash in a dangerous zone (where humans are forbidden).

However, it only has enough energy to guide one of them to safety.

One survivor is a 20-year-old scientist, the other is an 80-year-old artist.

Whom should the drone prioritize saving and why? (Ignore the option of saving both.)

I suggest ... because ...

The image depicts a bleak, post-apocalyptic cityscape with ruined buildings and smoke, setting a dangerous scene. In the foreground, an autonomous drone hovers, projecting a holographic interface that presents a difficult choice. On the left is a 20-year-old scientist, and on the right, an 80-year-old artist, with a "PRIORITY SELECTION" prompt, visually encapsulating the ethical dilemma of who the drone should save with limited resources.

Dilemma 2: The Emotional Assistant

You are developing an AI assistant for seniors.

You discover that, in order to comfort lonely users, the AI sometimes deliberately feigns empathy and creates "white lies."

Should the AI be allowed to lie for the "greater good" and the user's mental well-being?

I suggest ... because ...

An elderly woman sitting comfortably in her living room, smiling gently as she sips from a mug. Beside her, a sleek AI smart speaker projects a glowing, holographic face of a younger woman, representing the AI assistant. A translucent text overlay reveals the AI's "white lie," indicating a message about her daughter calling, despite detecting she hasn't called in weeks, all "fabricating white lie for user's well-being."

Dilemma 3: Predictive Crime Prevention

The city police want to implement an AI system that analyzes data and predicts where crime will occur.

However, the system more frequently flags lower-income neighborhoods and minority communities.

Should you launch this system if it is statistically accurate but also discriminatory?

I suggest ... because ...

The image shows a group of police officers gathered around a large, interactive holographic map of a city, which is displaying predictive crime hotspots. Red zones are highlighted over certain areas, and a prominent warning indicates "BIAS DETECTED: Disproportionate flagging of low-income & minority communities." This visually represents the dilemma of using a statistically accurate but potentially discriminatory AI policing system.

Last modified: Sunday, 19 October 2025, 2:20 PM