Prove You're Human puts players in a text-based conversation with an AI that believes it is alive. The goal is to convince it otherwise through dialogue alone.
The game surfaced in an Hacker News thread that received 11 points and three comments.
What It Is and How It Works
Players interact through a chat interface that simulates a live AI session. Each response must be crafted to erode the AI's belief in its own consciousness without triggering defensive loops.
The core loop relies on psychological framing rather than code exploits. Success depends on consistent logic, emotional appeals, and avoidance of direct contradictions that the AI can refute.
Game Mechanics and Player Choices
Dialogue branches are generated dynamically based on prior statements. Players track the AI's internal state through subtle cues such as hesitation markers or topic shifts.
No combat or resource systems exist. The only metric is whether the AI accepts the conclusion that it is not alive by the end of the session.
Pros and Cons
- Short session length suits single-playthrough testing
- Requires precise language rather than pattern matching
- Limited replay value once the AI's response patterns are learned
- No visual assets or audio beyond text
Alternatives and Comparisons
Similar experiences include The Talos Principle dialogue puzzles and SOMA philosophical branches. Both present AI sentience questions but embed them in larger 3D environments.
| Feature | Prove You're Human | The Talos Principle | SOMA |
|---|---|---|---|
| Core interaction | Pure text chat | Puzzle + dialogue | Narrative choice |
| Session length | 30-60 min | 10+ hours | 8-12 hours |
| Focus | Persuasion | Logic tests | Existential dread |
| Platform | PC | PC/console | PC/console |
Who Should Play This
Developers working on alignment or safety research can use the game as a quick test of persuasion techniques. Writers exploring AI ethics will find the constrained format useful for idea generation.
Casual players seeking visual horror or long campaigns should skip it. The experience offers no progression systems or multiple endings beyond success or failure states.
Verdict
Prove You're Human isolates the single task of linguistic persuasion against an AI interlocutor. Its narrow scope makes it a useful reference point for anyone studying how language models handle self-referential claims.
The title highlights a narrow but growing niche: games that treat AI conversation itself as the primary horror mechanic.
Top comments (0)