Google filed 2 patents covering 2 categories: AI & Machine Learning (1) and Networking (1). The AI & Machine Learning patent describes a coaching system for video games that uses chain-of-thought reasoning to explain optimal moves through natural language conversations based on gameplay footage.
The Networking patent covers uplink scheduling techniques for 5G networks that dynamically adjust PUSCH resources to reduce jitter and latency in extended reality and cloud gaming applications.
Google's single AI and machine learning patent tackles a longstanding problem in game coaching tools: the lack of transparency in AI recommendations. The system processes actual gameplay video or images and engages players in natural language dialogue, walking them through the strategic reasoning behind suggested moves rather than simply telling them what to do. Players can ask follow-up questions and probe the AI's logic interactively, transforming opaque machine learning outputs into educational conversations that explain the "why" behind optimal plays.
The Networking patent addresses the wireless infrastructure challenges of extended reality and cloud gaming, where traditional 5G uplink scheduling struggles to balance low latency with resource flexibility. Google's approach allows user equipment to modify PUSCH transmission occasions on a per-traffic-period basis within an already-configured grant framework, adapting uplink resources in real time without the delays inherent in requesting dynamic grants. This method delivers the responsiveness needed for wireless VR, AR, and game Streaming by letting devices adjust their transmission patterns as traffic demands shift, reducing both jitter and latency without sacrificing the efficiency gains of pre-configured resource allocation.
All data sourced from USPTO patent filings. Google Patents may take several weeks to index recent publications. If a link is unavailable, search for the patent number at USPTO Patent Public Search.