Neural Court: Ethics, Augmentation & Transformers

single-file interactive HTML game

Defend the human signal.

You are the senior analyst inside a neon tribunal where careless AI systems are trying to pass as judgment, evidence, friendship, and truth. Win by using the course principles: augmentation over automation, strong prompting, transformer basics, and the six ethical x_lens.

W1D1GPT, transformers, augmentation, human checking.

W1D2Prompt craft, real sources, anti-slop.

W1D3Rights, justice, utility, care, virtue, common good.

Transformer Engine

Repair the model’s attention core. A transformer is not magic: it predicts using patterns, context, and relationships between words.

Decode GPT

What does GPT stand for?

Attention Core

The transformer’s key advantage in language is that it can...

Prediction Trap

A confident AI answer should be treated as...

Mini System Log

SYSTEM: Transformer hum detected.
GOAL: Separate prediction from judgment.
STATUS: Awaiting human signal.

Augmentation Forge

Forge the correct workflow. The course emphasis is not “AI replaces you”; it is “human + AI with judgment, checking, and originality.”

Choose the workflow

Which one best represents augmentation?

Humanity Marks

Click the item that counts as “human signal.”

Empathy Filter

Why can emotional AI feel risky?

Forge Meter

HUMAN INPUT: source, voice, context.
AI OUTPUT: structure, draft, options.
FINAL POWER: edited judgment.

Work & Prompt Lab

Defeat the Slop Machine by choosing concrete constraints: role, task, sources, layout, data, and output format.

Prompt Duel

Pick the strongest prompt.

Graph Lock

Why should you give real data before asking AI for a graph?

Sort the Prompt Parts

Click a token, then click the correct slot.

“You are a legal analyst”

“Use these article links/data”

“Output one HTML file”

ROLE

SOURCES

FORMAT

Anti-Slop Console

PROMPT QUALITY = role + task + constraints + evidence + output.
VAGUE ADJECTIVE DETECTED: “interesting”.
ACTION: replace with concrete mechanics.

Ethics Court

Final boss: assign the right ethical lens to each courtroom question. Ethics asks who is harmed, who benefits, who is responsible, and what kind of people/society we become.

Case 1: A school AI flags a student for cheating. The evidence is unclear, but the punishment is serious. Which lens asks whether the student has appeal, explanation, and human review?

Select a lens to argue the case.

verdict rendered