Decentralized Artificial Intelligence through Controlled Emergence (DICE)

The Decentralized Artificial Intelligence through Controlled Emergence (DICE) program seeks to develop the theory and algorithms for decentralized coordination and local inference control to enable a scalable, adaptive, and resilient collective of heterogeneous AI agents. The program aims to harness the scalability and adaptability of self-organizing systems while minimizing risks and ensuring that the collective behavior remains predictable and aligned with intended outcomes. Key aspects of the program include: • Developing decentralized AI systems that can autonomously execute sustained long-time-horizon missions in contested environments • Ensuring that the collective behavior of the AI agents remains under human control • Minimizing risks and ensuring predictability of the AI system's behavior • Developing a scalable and adaptive AI system that can operate in dynamic environments • Ensuring that the AI system is resilient and can recover from failures or disruptions