Explore a groundbreaking conference talk on ADR-X, an innovative ANN-based contextual multi-armed bandit rate adaptation technique for wireless gaming devices. Delve into the challenges of rapid channel variations in console gaming and learn how ADR-X achieves 10× lower packet losses than existing schemes while running 100× faster than state-of-the-art reinforcement learning approaches. Discover how this technique leverages communication theory and domain knowledge to create an efficient ANN suitable for deployment on embedded gaming devices. Gain insights into the extensive evaluations and user studies that revealed the limitations of current rate adaptation schemes and the need for significantly lower packet loss rates to ensure high-quality gaming experiences.