A Head-to-Head Battle of AI Models in Game Development

In this thrilling session, we put four (maybe five) leading AI models—DeepSeek, ChatGPT, Claude, Gemini, and Grok—to the ultimate test: building a fully playable game from scratch!

Using a single game concept, each AI will be challenged to generate game mechanics, logic, visuals, and playability with minimal human intervention. We’ll compare their performance across key metrics, including speed, completeness, code quality, and overall effort required to get a polished, working game.

Attendees will witness firsthand how different AI models tackle the same challenge, where they excel, and where they fall short. By the end of this session, we’ll crown the most effective AI for game development—and lay the foundation for an AI benchmarking system for software engineering tasks.

Whether you’re a developer, AI enthusiast, or game designer, this is your chance to see AI battle it out in the ultimate game dev showdown!

Key Takeaways:

✅ Which AI model is the best at generating a complete, playable game?
✅ How much human effort is needed to refine AI-generated code?
✅ Does AI understand game mechanics, physics, and user experience?
✅ How do different models handle debugging and iteration?
✅ Insights into the future of AI-assisted game development