Google is redesigning its classic search page to make LLM-driven search the primary experience for users. Announced at last week’s I/O conference, the traditional single-line query box will be replaced by an “intelligent search box” that can, when useful, move users into a back-and-forth conversation with Gemini. Rather than returning the long-established list of blue links, Search will increasingly surface LLM-generated answers and encourage follow-up questions so users can clarify and narrow their intent. Traditional link-based results won’t vanish entirely but will be deprioritized compared with responses produced by language models.
The update brings plain search closer to features that have already been appearing over the past few years, such as AI Overviews—Gemini summaries of web content—and AI Mode, which lets people chat with a conversational assistant to explore answers. Google also plans to introduce LLM-powered agents that continuously monitor web pages and alert users when relevant updates occur. This expands on the concept behind Google Alerts: instead of manually checking results, users can define criteria and let an “information agent” watch for changes like price shifts or breaking developments, then send a consolidated update when conditions are met.
Liz Reid, Google’s head of search, explained that users could, for example, have an agent monitor market movements within tightly defined parameters; the agent would then determine which datasets and tools to use, watch those sources, and notify the user with a summarized report and links for deeper reading. According to Google, AI Overviews are invoked more than 2.5 billion times each month, and AI Mode has roughly one billion monthly users. For context, Google handles about 13.7 billion searches every day, meaning only about 0.85% of the search audience uses these LLM features each month.
The Gemini-powered interface will start to roll out this week, while the agent-driven monitoring features—powered by technology Google calls Antigravity—are slated for later in the year. Sundar Pichai, Google’s CEO, said the company invests in cutting-edge models so it can make them widely available. The company plans to offer the redesigned search box and the generative interface at no cost to users, while background monitoring and notification features will be reserved for Google AI Pro and Ultra subscribers later this year. Google also noted that Gemini Spark, a personal assistant currently in beta for select Ultra subscribers in the U.S., will move toward production soon.
For marketers who depend heavily on conventional SEO, these changes signal a shift worth watching. Google’s recent moves appear driven by three aims: keep users within Google’s ecosystem, provide answers without always surfacing external links, and increasingly rely on LLMs to interpret and present web content. While other AI providers have shifted toward token-based pricing to cover model expenses, Google’s broader revenue streams allow it to subsidize advanced models more aggressively. That advantage makes it likely Google’s new LLM-first search features will deepen over time and become a lasting element of its product offering.
Sources url – https://www.marketingtechnews.net/news/google-search-optimisation-pushes-traditional-search-results-further-down-the-page/
