Google introduces Web Guide within Search Labs, a new feature that groups search results based on AI analysis. The feature uses a customized version of Gemini for this purpose.
Google tests a new feature, Web Guide, which ranks search results using AI to make information more easily accessible. Web Guide utilizes a “customized version of Gemini to better understand queries and web content”, according to Google. The experiment is currently available to users within Search Labs. It is uncertain whether this experimental feature will be implemented.
AI groups web pages
Web Guide groups links to websites based on specific aspects of a search query. Related pages about a topic appear grouped together. The technology behind the feature is a customized version of Gemini, Google’s AI model. This model analyzes both the query and relevant web content to bring forward the most meaningful results.
The feature uses a technique where multiple related queries are executed in parallel, also known as fan-out-search. This increases the chance of displaying pages that might otherwise go unnoticed. Web Guide is particularly intended for open or complex queries, such as “how to solo travel in Japan” or “tools to stay in touch with family across different time zones.”
Web tab
Users who sign up for Search Labs can activate Web Guide via the Web tab in the search results. They can return to the classic search results view at any time. Google plans to display AI-organized results in other areas of the search engine in the future, such as in the “All” tab. This will happen gradually, depending on user utility.
read also
OpenAI working on AI web browser and challenges Google Chrome
Web Guide is currently an experimental feature within Google Search Labs. The project is part of Google’s broader efforts to use AI to enhance search experiences.