This project implements an experimental game-based approach to studying human-AI interaction by inverting the traditional Turing Test paradigm. Instead of testing AI's ability to appear human, this study examines how humans modify their behavior when attempting to blend in with AI agents.
This implementation addresses several key research questions:
- How do humans modify their communication patterns when deliberately attempting to "pass" as AI?
- What behavioral markers do AI systems use to distinguish human from AI-generated responses?
- What strategies do humans employ to emulate AI-like communication?
The study employs a game-based methodology where:
- Participants engage in role-played conversations
- AI systems generate contextually appropriate responses
- Pattern analysis algorithms track communication markers
- The system attempts to identify the human participant
- Limited to text-based interactions
- Single-player implementation
- Fixed set of historical figures
- Basic pattern analysis metrics
Planned enhancements include:
- Expanded pattern analysis capabilities
- Additional historical figures
- Multi-player support
- More sophisticated detection algorithms
The enitre project is inspired by https://www.youtube.com/watch?v=MxTWLm9vT_o