Abstract:Powerful artificial intelligence systems are often used in settings where they must interact with agents that are computationally much weaker, for example when they work alongside humans or operate in complex environments where some tasks are handled by algorithms, heuristics, or other entities of varying computational power. For AI agents to successfully interact in these settings, however, achieving superhuman performance alone is not sufficient; they also need to account for suboptimal actions or idiosyncratic style from their less-skilled counterparts. We propose a formal evaluation framework for assessing the compatibility of near-optimal AI with interaction partners who may have much lower levels of skill; we use popular collaborative chess variants as model systems to study and develop AI agents that can successfully interact with lower-skill entities. Traditional chess engines designed to output near-optimal moves prove to be inadequate partners when paired with engines of various lower skill levels in this domain, as they are not designed to consider the presence of other agents. We contribute three methodologies to explicitly create skill-compatible AI agents in complex decision-making settings, and two chess game frameworks designed to foster collaboration between powerful AI agents and less-skilled partners. On these frameworks, our agents outperform state-of-the-art chess AI (based on AlphaZero) despite being weaker in conventional chess, demonstrating that skill-compatibility is a tangible trait that is qualitatively and measurably distinct from raw performance. Our evaluations further explore and clarify the mechanisms by which our agents achieve skill-compatibility.

Design of Game Circuit System Based on Horn Chess

Design and Implementation of Chinese Chess Based on Manipulator

Design and Implementation of Intelligent Chinese Chess System Device

Design of a speech interaction system for man-machine confrontation

Demo: ROX Player A Xiangqi Playing Robotic System.

Design and implementation of a PCB testing system

The Circuit Experiment Teaching Based on CDIO/DEC

Design and improvement of the pruning algorithm of the Chinese chess in the computer games

Design and Implementation of Network Chinese Chess Game with Hall

THE GAME TREE SEARCH ALGORITHM BASED ON OPTIMIZED ITERATIVE DEEPENING

The application of artificial intelligence in board games

Design of Digital Signal Generator Based on DSP and SOPC

The Design of Control System Based on ARM and High-speed Pulse Generator

Design of Multi-Axis Motion Controller and Its Application on Manipulators

Computer Aided Design of Arm and Hand of Chess Playing Robot

AI in Human-computer Gaming: Techniques, Challenges and Opportunities

Designing Skill-Compatible AI: Methodologies and Frameworks in Chess

Design of Interactive Teaching System of Physical Training Based on Artificial Intelligence

Development of a Neuromorphic Computing System

Application of Q-Learning and RBF Network in Chinese Chess Game System

Neuromorphic Co-Design as a Game