WE'VE BEEN PUT IN THE VIBE SPACE

Summary from the TLDR newsletter

Search and recommendation have traditionally been kept separate, because while they occupy opposite ends of the same spectrum conceptually, they have different tooling, architectures, and machine learning goals. Large language models introduce expectations that need to be managed carefully because it is hard to guide customer intent in an open-ended solution space. This may be frustrating for consumers until specific LLM use cases are sorted out and made less general.