Microsoft researcher builds a working neural network in Age of Empires II using goats

From Pixel to Protocol: How an AI Scientist is Rewriting the Rules of History

In the cutting-edge world of artificial intelligence, some breakthroughs happen in sterile labs and others are sparked by a nostalgic passion. Meet Adrian de Wynter, a researcher at Microsoft and a scholar at the University of York who bridges these two worlds with remarkable skill. While he delves into the fundamental challenges of AI and related technologies, his unique perspective is deeply rooted in a hobby many associate with pure entertainment: the strategic world of gaming.

De Wynter’s journey into advanced computation is matched only by his lifelong dedication to gaming. Since 1999, he has been immersed in the complex, deep-strategic mechanics of Age of Empires II. This history isn’t just a footnote; it forms the intuitive foundation upon which his contemporary research is built.

The true fascination lies where these two disparate fields intersect. De Wynter is not simply applying AI to abstract problems; he is using the historical context and complex decision-making structures of strategy games to explore how neural networks can understand, predict, and even generate human history.

In a recent study, de Wynter explored the dynamic relationship between historical strategy and advanced machine learning. The research focused on leveraging neural network architectures to analyze the intricate systems and emergent strategies found within classic historical simulations.

This work moves beyond simple pattern recognition. It investigates how complex historical timelines, populated by competing civilizations and evolving economies, can be modeled effectively by modern AI systems. By studying these game mechanics—resource management, military expansion, technological progression—researchers gain powerful new insights into the underlying logic of historical causality.

The experiment demonstrated the potential for neural networks to grasp not just immediate tactical decisions but also long-term, high-level strategic planning that defines grand historical narratives. It suggests that the principles governing empire building and societal evolution possess a mathematical structure ripe for AI dissection.

Adrian de Wynter’s research is thus illuminating a unique pathway: turning the lessons of complex gaming history into sophisticated tools for advancing artificial intelligence. It proves that the deep, emergent strategies found in games like Age of Empires II hold valuable keys to understanding the complexity of human civilization itself.

Buy on Amazon