This project is an interactive game where users engage in conversations with AI agents posing as historical figures. The goal is for players to identify which participant is human, effectively conducting a reverse Turing Test.
- Multiple AI agents and one human user
- Automated conversation system with AI agents
- Structured question-and-answer phases
- Voting system to guess the human
- Optimized API usage: The API token usage is optimized by utilizing
gpt-4o-mini
for most of the game, except during the voting result computation where higher precision is needed.
- Backend: Flask (Python)
- Frontend: React
- AI: OpenAI GPT API
-
Clone the repository:
git clone [repository_url] cd reverse-turing-test-game
-
Backend setup:
cd backend pip install -r requirements.txt
-
Frontend setup:
cd frontend npm install
-
Set OpenAI API key: Place your API key in the
backend/keys/openai_api_key
file. -
Start the backend server:
cd backend python app.py
-
Start the frontend:
cd frontend npm start
-
Access the game via
http://localhost:3000
.
- The game starts with 3 AI agents (posing as historical figures) and 1 human player.
- Players take turns asking and answering questions.
- After all questions, participants vote on who they think is human.
- The winner is determined by the voting results.
Feel free to open issues or submit pull requests if you'd like to contribute. Contributions are welcome!
This project is licensed under the MIT License. See the LICENSE
file for more details.