YouTube music uses large language models to explain new artist picks
Explainable Recommendations at Scale: LLM Rationales for YouTube Music Artist Discovery
Artificial IntelligenceInformation Retrieval
Summary
Music streaming services want to help people discover new artists, but listeners often stick to familiar favorites. The authors show how YouTube Music created explanations in everyday language to explain why new artists are recommended. Because running large language models in real-time is too slow and costly, they pre-generate these explanations ahead of time. Tests on millions of users proved this approach encourages people to explore more new music while keeping the service fast and responsive.
What this means in practice
- •For music streaming engineers: Build scalable recommendation systems that precompute natural language explanations for new content to boost user exploration without slowing service response times.
- •For e-commerce developers: Integrate offline large language model generated explanations to improve user trust and discovery of less-known products in recommendation engines.
Authors
Xiao Liu, Yanwei Song, Srivaths Ranganathan, Yuan Chen, Zheyun Feng, Parker Steenburgh, Jochen Klingenhoefer, Nathan Lasche, Gergo Varady, Tim Steele
Abstract
Modern music streaming platforms face a persistent tradeoff: exploiting familiar content versus driving the exploration of novel items. While users frequently desire discovery, they hesitate to select unknown artists over proven favorites. Providing transparent, natural language rationales that explain why an unexplored item is recommended lowers this barrier. However, while Large Language Models (LLMs) excel at this nuanced explainability, their real-time deployment is severely bottlenecked by prohibitive inference costs and computational overhead. In this paper, we present an industry case study of a decoupled recommendation architecture that successfully scales exploration without compromising latency. Our system isolates LLM inference asynchronously offline, pre-computing personalized candidate pools of undiscovered artists alongside tailored rationales. Large-scale online A/B experiments validate our design. We demonstrate that combining LLM-backed recommendations with these explanatory rationales significantly reduces the trust barrier for new content, yielding statistically significant improvements in both user exploration and overall engagement on the discovery surfaces.