Bing Generative Search: Microsoft’s New AI-Driven Experience
Microsoft is exploring a new approach to search with its Bing generative search. This feature combines Bing’s search results with large and small language models (LLMs and SLMs) to generate AI-driven search results. This innovation aims to provide a more effective and user-friendly search experience.
How Bing Generative Search Works
Bing generative search integrates traditional search results with AI-generated content. It understands search queries, reviews many information sources, and generates search results in a new layout. This layout includes document indexes, answer summaries, and traditional search results. The AI-driven results fulfill user queries more effectively, providing concise answers and relevant links.
Features and Benefits
A notable feature is the AI-generated search results page. For example, a search for “What is a spaghetti western?” not only provides traditional search results but also includes an AI-generated overview of the film subgenre, its history, and top examples. This layout is easy to read and navigate, enhancing the user experience by offering comprehensive answers and sources for deeper exploration.
Current Testing and User Impact
Microsoft is currently testing this feature with a small set of users. The company aims to gather feedback, refine the experience, and ensure it benefits users before a broader rollout. Microsoft acknowledges the potential impact on web traffic and emphasizes the need to monitor click-through rates and maintain a healthy web ecosystem.
Comparison to Google’s AI Overviews
Bing’s generative search is similar to Google’s AI Overviews, which also uses AI to create summaries for search queries. Both aim to improve search efficiency but have faced concerns about potentially reducing organic traffic to websites. Microsoft is committed to closely monitoring these impacts and adjusting the feature to support web traffic and ad revenue for publishers.
Future Prospects
As Bing continues to integrate AI into its search functions, it is expected to provide users with more dynamic and useful search results. The ongoing testing phase will be crucial in refining the feature and ensuring it meets user needs while maintaining web traffic to publishers.