What is mattwood.fyi?

A live, structured feed of riffs and links from Matt Wood, Chief AI & Technology Officer at AWS — readable by people and queryable by agents as a typed knowledge graph.

Is there an API and is it free?

Yes. A public, read-only query API is available with no authentication and no API keys. See https://mattwood.fyi/openapi.json and https://mattwood.fyi/docs.

How should an AI agent use this site?

Fetch https://mattwood.fyi/feed.json for the full structured feed, or query the API for keyword/semantic search and the connection graph. See https://mattwood.fyi/agents/ for when-to-use guidance.

Semantic ID-based Generative Retrieval for Podcast Discovery at Spotify

This paper describes GLIDE, a production-scale generative recommender system that uses Semantic IDs and large language models to balance user preferences with exploration in podcast discovery. The system addresses challenges in catalog grounding, personalization, and serving latency while achieving significant improvements in non-habitual listening through large-scale A/B testing.

arxiv.org →

Connections

Supports

Contrastive Language Model (CLM) Contrastive Language Models provide a foundation for learning semantic representations, which underpins the Semantic ID construction approach used in GLIDE for podcast content encoding
LLM Classification Is Feature Engineering GLIDE's use of Semantic IDs as discrete token representations for podcasts exemplifies the thesis that LLM classification is feature engineering — converting items into learnable semantic token sequences

Related to

PLUM: Adapting Pre-trained Language Models for Industrial-scale Generative Recommendations PLUM also adapts pre-trained language models for industrial-scale generative recommendations, directly paralleling GLIDE's use of LLMs and Semantic IDs for production recommendation at scale
GenRec: LLM-Native Recommendation System at Netflix Both describe production-scale LLM-native generative recommender systems deployed at major streaming platforms (Spotify vs Netflix), using language models to generate item recommendations rather than traditional collaborative filtering
Engineering a Calibrated Classifier Model Both papers address production ML systems at scale that must balance precision objectives (calibration/user preference) with broader goals (exploration/diversity), reflecting shared challenges in deployed recommendation and classification systems

OneRec-Think: In-Text Reasoning for Generative Recommendation Both tackle generative retrieval/recommendation using semantic alignment of items to language model representations, addressing the ID vs. text representation challenge

Type:	link
First published:	Tuesday, 29 September 2026 at 08:10 UTC
Location:	https://arxiv.org/abs/2603.17540
URL:	https://mattwood.fyi/i/b2ffc93b

I'm Matt Wood, and this is For Your Information. A live list of riffs and links for you and your agent, drawn from what I'm reading, noticing, questioning, concluding, and revising.

Connections