Bing is testing a new generative search experience that integrates AI and large language models (LLMs) to create AI-generated search results pages. The experiecne is very similar to Google's AI Overviews. This initiative builds on the LLM-powered chat answers introduced last year. The new search experience aims to provide dynamic, bespoke responses to user queries by combining traditional search results with AI-generated content.
For instance, a search for "What is a spaghetti western?" will yield an AI-generated explanation covering the film subgenre's history, origins, and notable examples, along with links and sources for further exploration. Regular search results remain prominently displayed.
The generative search experience leverages both large and small language models (LLMs and SLMs) to understand queries, review millions of sources, and generate relevant content dynamically. This approach aims to fulfill user intent more effectively while maintaining traditional search result features and increasing clickable links.
Bing has optimized its methods to ensure accuracy and is monitoring the impact of generative search on web traffic. Early data suggests that this new experience maintains website clicks and supports a healthy web ecosystem. The rollout is gradual, with Bing seeking user feedback to refine the experience before broader availability.
Users can provide feedback using thumbs up and thumbs down icons or the Feedback icon on the search results page.