tldr - powered by Generative AI

• Spotify is building an encoder system that can encode a user's state into embeddings that are sensitive to actions and changes in satisfaction.
• They have built a simulator that can simulate user reactions to certain content, which is used to train the recommendation algorithm.
• A/B testing is used to compare agents trained on good simulators versus less good simulators.
• Spotify is transitioning to a more farsighted approach to content recommendation, optimizing for long-term fulfilling content diet rather than clicks or streams.
• They are investing in data instrumentation to understand how users interact with their content and to create reusable data sets.
• Spotify has shared machine learning models that provide information on user affinities, similarities, and clustering, which are useful for many different features.
• Machine learning models are created for specific use cases, such as Discover Weekly or search, and optimize for different goals.

tldr - powered by Generative AI

• Netflix faces a variety of challenges in personalization, including improving diversity, freshness, and fairness of recommendations
• Recent trends in improving personalization include using deep learning for recommendations, causality to understand the impact of recommendations, bandits to optimize recommendations, and objectives to measure success
• Deep learning involves learning embeddings for user and item IDs and using them to make predictions
• Causality involves understanding the impact of recommendations on user behavior using randomized experiments
• Bandits involve optimizing recommendations by balancing exploration and exploitation
• Objectives involve measuring success using metrics that capture user satisfaction and fairness
• Netflix is hiring for research and summer intern positions in personalization