Google has introduced AI Overviews in its search results for users in the United States, with plans to expand this feature globally. The AI Overviews, which are part of the Search Generative Experience (SGE), aim to provide more comprehensive answers to complex queries by leveraging generative AI technology powered by Google Gemini.
AI Search Capabilities
AI Overviews are designed to offer detailed responses by summarizing content from various web sources. They are particularly targeted at complex questions where traditional search results may fall short. According to Hema Budaraju, Senior Director of Product for the Search Generative Experience at Google, these AI-generated summaries result in higher click-through rates compared to standard web search results. However, Google will not separate click data for AI Overview links in its Search Console, making it difficult for content creators to gauge the performance of their content within these overviews.
New Features and Limitations
In addition to AI Overviews, Google is rolling out several new AI features through its Search Labs initiative. These include multistep reasoning capabilities, planning tools, AI-organized search results, and a new feature in Google Lens that allows users to ask questions about videos. Despite these advancements, AI Overviews will not be displayed for all queries. They are reserved for instances where Google believes the AI can add significant value beyond what is already provided by traditional search results. The company has not disclosed the specific percentage of queries that will trigger an AI Overview.
Google’s decision to integrate AI Overviews without distinguishing their data in Search Console has raised concerns among content creators. Some speculate that this approach may obscure the true impact of AI-generated answers on click-through rates and overall traffic to publishers’ sites. Additionally, Google assures that ads will continue to appear in their designated slots on the search results page, regardless of the presence of AI Overviews. The company emphasizes that the introduction of AI Overviews is not intended to replace featured snippets, which will continue to be part of the search results.
Development and Integration of AI Technologies
Google is starting to roll out AI Overviews to users in the US and soon around the world. Liz Reid, Google’s newly installed head of Search, has been working on all parts of AI search for the last few years. Reid mentions that AI Overviews are meant to give a general sense of the answer to a query along with links to resources for more information. Google announced these AI features publicly at its I/O developer conference.
Google is using its Gemini AI to figure out what users are asking about, whether they are typing, speaking, taking a picture, or shooting a video. A new specialized Gemini model is used to summarize the web and show an answer. Gemini is also used to design and populate the results page. Reid highlights that not every search will get AI Overviews, particularly simple navigational queries. Reid provides an example of local search where Gemini can combine information across the Knowledge Graph and the web. The combination of the Knowledge Graph and AI is key for Reid and her team.
Balancing Accuracy and Creativity
Reid emphasizes that Gemini’s job is to ensure users get accurate data, even if they ask questions in unconventional ways. Reid notes that AI Overviews may not be fun or charming but are skewed toward factuality. Reid acknowledges the challenge of balancing creativity and factuality in language models. Reid mentions that early data shows this new way of searching will actually lead to more clicks to the open web. Reid observes that young users are particularly interested in a human perspective on their queries.
Last Updated on November 7, 2024 8:25 pm CET