ABOUT

SECRET CODER

Secret Coder is a short, browser-based game about what happens to meaning when AI systems get involved in moving language around.

You play as a Secret Coder: someone whose job is not to decide whether a piece of content is true or false, but to trace where it came from, notice how it was transformed, and protect a reference to what the original speaker actually meant — before the content travels any further.

One playthrough. About 3–5 minutes. A new random scenario every time.

THE CENTRAL QUESTION

When AI systems transform language, how can we preserve a reference to what the original speaker actually meant?

Secret Coder is inspired by the Secret Code concept and the speculative Right to Semantic Integrity.

HOW IT WORKS
See a piece of AI-generated or AI-transformed "slop."
↓
Inspect its provenance — the systems it passed through.
↓
Compare the original statement against the current version.
↓
Diagnose what kind of semantic drift occurred.
↓
Choose a Secret Code naming the vulnerability.
↓
Attach a semantic anchor: the meaning that must survive.
↓
Watch the content transform again, further from the source.
↓
Check whether your anchor's meaning survived the trip.
FOUR IDEAS THE GAME PLAYS WITH
PROVENANCE

Where did it come from?

TRANSFORMATION

What happened to it?

TRANSPARENCY

Can we tell AI was involved?

SEMANTIC INTEGRITY

Does it still mean what the originating speaker intended?

Provenance can tell us where content came from. Transparency can tell us AI was involved. Neither one tells us whether the original meaning survived — that's what semantic integrity asks, and what a semantic anchor is designed to protect.

A Secret Code isn't a fact-check label. It's a speculative semantic layer — a persistent reference attached to content, so that future transformations can be checked against it.

PLAY SECRET CODER