About Truth or Dare AI

Inspiration

The classic party game Truth or Dare has been a staple of social gatherings for generations. However, we noticed that traditional versions often have limitations:

• Static question banks that quickly become repetitive
• Questions that might be inappropriate for certain groups
• Lack of personalization based on player preferences
• Limited adaptability to different social settings

We saw an opportunity to modernize this beloved game using AI technology to create a more dynamic, inclusive, and engaging experience.

What We Built

Truth or Dare AI is a web application that brings the classic party game into the digital age. Using Streamlit for the frontend and Google's Gemini Pro API for AI-powered content generation, we created a game that:

• Adapts questions based on player demographics
• Offers multiple intensity levels (mild, funny, spicy, drunk)
• Generates personalized content in real-time
• Maintains a fun and appropriate atmosphere for all players

How We Built It

The project was built using:

• Frontend: Streamlit for a responsive, user-friendly interface
• Backend: Python for game logic and state management
• AI Integration: Google's Gemini Pro API for dynamic content generation
• Development Tools: Git for version control and VS Code for development

The application features a carefully curated question bank with over 100 questions and dares, categorized by intensity and demographics. We implemented session state management to maintain game progress and player preferences.

Challenges We Faced

1. Content Curation: Creating a diverse and appropriate question bank that appeals to different audiences while maintaining a fun atmosphere
2. AI Integration: Fine-tuning the Gemini Pro API to generate contextually appropriate and engaging content
3. State Management: Implementing effective session state management in Streamlit to maintain game flow
4. User Experience: Balancing simplicity with feature richness to create an intuitive interface

What We Learned

• The importance of user experience in game design
• Effective state management in Streamlit applications
• Best practices for AI prompt engineering
• Creating inclusive content for diverse audiences
• Balancing different intensity levels in social games

Future Improvements

We plan to:

• Add multiplayer support with real-time interaction
• Implement a scoring system
• Expand the question bank with user-generated content
• Add themed question packs for specific occasions
• Incorporate machine learning to better personalize questions based on player responses

Built With

• git
• python
• streamlit
• venv
• vscode