Google is fundamentally redesigning its search engine, transitioning from a system that delivered lists of links to one that generates conversational answers, autonomous agents, and interactive interfaces. This transformation, which company executives described as the most significant shift in the service’s history, could further reduce web traffic for publishers who rely on search referrals.
The update was announced during Google’s annual developer conference held in Mountain View, California, on May 14. The company stated that the changes will roll out to users in the United States starting this week, with a global expansion planned over the coming months.
AI Overviews replace traditional search results
At the center of the redesign is a feature called AI Overviews, which will appear at the top of search results pages. Instead of presenting a list of blue links, the system will generate a paragraph-length answer to a user’s query, created by an artificial intelligence model. Google said this feature is designed to provide faster and more complete answers to complex questions.
The AI Overviews are powered by a customized version of Google’s Gemini language model. According to the company, the system can handle multi-step reasoning, such as planning a trip itinerary or comparing products across multiple categories. The feature currently supports text-based questions, but Google plans to add image and video inputs in future updates.
Liz Reid, Google’s vice president of search, said during a presentation that the goal is to make search “do the work for you.” She stated that the system can now handle tasks that previously required users to visit multiple websites and synthesize information themselves.
Impact on publisher traffic
The transition to AI-generated answers has raised concerns among publishers and content creators. Many news organizations and information websites derive a significant portion of their traffic from Google Search. If users receive answers directly on the search results page, they may have less reason to click through to external sites.
Google acknowledged this concern in a blog post published alongside the announcement. The company said it will continue to prioritize “high-quality content” and will include links to sources within the AI Overviews. However, early tests of the feature showed that in some cases, the AI summary took up nearly the entire screen on mobile devices, pushing traditional search results and links further down the page.
Similar changes at other technology companies have already demonstrated the potential effect. Social media platform Reddit and information sharing site Wikipedia have both reported changes in referral traffic patterns following the introduction of AI-powered summaries in search products.
Competitive context
Google’s announcement comes amid intensifying competition in the AI search space. Microsoft’s Bing search engine, which integrated OpenAI’s GPT-4 language model in February 2023, has seen a modest increase in market share. Several startup companies including Perplexity AI have also launched AI-native search products that generate direct answers instead of link lists.
Industry analysts have noted that Google’s position as the dominant search engine, commanding more than 90 percent of the global search market, means that even small changes to its product can have outsized effects on the web ecosystem. The company reported that it processes more than 8.5 billion searches per day.
Technical details and limitations
Google said the AI Overviews system will initially handle queries where the AI model has high confidence in its answer. For topics involving breaking news, health information, or financial data, the system will default to traditional search results or display prominent disclaimers. Reid stated that the company has implemented safeguards to reduce the risk of generating inaccurate or harmful responses.
The underlying Gemini model has been trained on a large corpus of text data, including publicly available web pages. Google representatives confirmed that the company continues to use web content for training, a practice that has drawn criticism from some publishers who argue that their material is being used without compensation.
Google has not disclosed the computational cost of generating AI Overviews for each query, but independent estimates suggest that AI-powered search requires significantly more processing power than traditional keyword matching.
Upcoming features and rollout timeline
Beyond AI Overviews, Google previewed additional search features scheduled for release later this year. One such feature will allow users to adjust the complexity of search results, choosing between simplified explanations and more detailed responses. Another tool will enable users to pose follow-up questions directly from the search results page without restarting the query.
Google also announced plans to integrate its AI search capabilities with other products including Google Maps and Google Shopping. Users searching for local businesses may see AI-generated summaries of reviews and recommendations, while product searches could display automatically generated comparison tables.
The company confirmed that AI Overviews will initially launch in English for users in the United States. Google plans to expand to additional languages and regions in stages throughout 2024 and 2025. The company has not specified a timeline for global availability.
Source: Delimiter
