I wanted to build something simple but technically meaningful. Tic Tac Toe looks small, but when you add AI, state tracking, and persistence, it becomes a proper logic system project.
Introduction
I built a fully working Tic Tac Toe browser game. I did not manually write every line. I used AI tools to speed up development, but I controlled the structure, features, and logic validation.
This project helped me understand how modern development is changing. AI can generate code fast, but correct system behavior still depends on developer understanding.
Game Features
- 2 Player Mode
- AI Mode (Easy / Medium / Hard)
- Hard Mode uses Minimax Algorithm
- Win Streak Tracking
- Badge Unlock System
- Match History Storage
- Local Storage Persistence
- Responsive UI
Tools I Used
I used multiple AI tools to speed up development:
- Claude → UI structure and layout ideas
- DeepSeek → Logic improvements and bug fixing
- Qwen → Code understanding and optimization help
My Actual Work:
- Deciding game architecture
- Designing feature logic
- Testing game behavior
- Fixing logical bugs
- Merging outputs into one working system
How I Used Prompting
Instead of asking AI to “make a game”, I gave structured system prompts.
Then I optimized performance using better prompts.
Technical Things I Learned
- Minimax decision logic
- Browser localStorage state handling
- DOM event driven programming
- Game state management
- Recursive performance optimization
Reality of Using AI in Development
AI helped me generate base structure fast. But working software required testing, debugging, and logic validation.
- AI gives structure
- Developer ensures correctness
- Testing is still manual
- System design thinking is still human
Future Improvements
- Online Multiplayer
- Leaderboard Database
- PWA Mobile Version
- Better AI Difficulty Scaling
Play the Game
Final Note
This project helped me understand practical AI-assisted development. AI speeds up coding, but system design and logic correctness still depend on the developer.