The internet’s most famous blank white page is getting its biggest makeover in 25 years. At its annual I/O developer conference on Tuesday, Google unveiled a sweeping transformation of its core search engine, converting the traditional one-line search box into an expansive, AI-driven hub. The company is fundamentally shifting from a tool that returns a list of web links to one that can independently browse the internet, draft code, and act on users' behalf.
The iconic slender search bar is officially expanding. Google is rolling out a redesigned box that dynamically grows to fit long, conversational questions rather than just short keywords. Powered globally by the new Gemini 3.5 Flash AI model, the upgraded box allows users to type complex queries, upload files, photos, videos, or even input Chrome tabs. Liz Reid, Google’s vice president of Search, noted during a briefing that the company wanted to eliminate the friction between automated AI summaries and its chatbot-like AI Mode.
According to a report by The Verge, Reid explained: 'We really work to make that much more seamless and simplified, so that for most users, they don’t have to think about where to go, they can just go to the search box they’re familiar with and it feels like they get the best experience afterwards.' Desktop and mobile users worldwide will now be able to ask follow-up questions directly from an AI Overview summary, shifting smoothly into a continuous back-and-forth conversation as the search engine retains the chat context.
AI agents that track listings and make phone calls
Google is also launching 'information agents' designed to complete long-term research tasks in the background 24/7. Instead of repeatedly checking websites for updates, a user can give an agent a detailed list of requirements, such as apartment-hunting parameters or tracking when a professional athlete drops a new sneaker collaboration. The agent will continuously monitor blogs, news sites, social media, and real-time data feeds, sending synthesized alerts when a match appears.
Furthermore, Google is expanding its automated booking capabilities for US users this summer. The system will search for real-time pricing and availability for local services, such as finding a specific type of private karaoke room on a Friday night, and provide direct booking links. For certain industries like home repair, pet care, and beauty, users can instruct Google to place phone calls to businesses on their behalf. This integration with Google Duplex technology, first demoed in 2018, now becomes a seamless part of the search experience, allowing users to delegate tasks without lifting a finger.
The move toward agentic search marks a significant departure from the traditional 'ten blue links' model. Google has been experimenting with proactive assistance for years, including Google Assistant and earlier versions of Duplex, but this is the first time such capabilities have been deeply embedded into the core search engine itself. The information agents are designed to learn user preferences over time, becoming more personalized as they handle recurring tasks like tracking price drops or monitoring local events.
Competitors are also moving in this direction. Microsoft’s Bing with Copilot has offered some agent-like features, and startups like Perplexity AI have gained traction by providing conversational search with citations. However, Google’s vast data infrastructure and existing user base give it a unique advantage. The company processes over 8.5 billion queries per day, and integrating agents into that flow could dramatically change how people interact with the web.
Instant coding and mini-apps on the fly
One of the most radical additions to Search is the ability to generate custom software interfaces on command. Using Google’s Antigravity platform and the coding capabilities of Gemini 3.5 Flash, the search engine can build custom, interactive visual tools from scratch to help answer complex topics. If a user asks about astrophysics or how a mechanism works, Search can code custom layouts featuring interactive graphs, tables, and real-time simulations.
For ongoing personal tasks, Search can create personalized dashboards or trackers that function like 'mini apps.' A user trying to establish a health routine can ask Google to design a custom fitness tracker. The search engine will generate software code on the fly, integrating live local maps, weather data, and reviews to keep the tool up to date week after week. This blurs the line between search and web application development, allowing anyone to create bespoke tools without coding knowledge.
The Antigravity platform, which has been in development for several years, leverages a lightweight runtime that can execute generated code in a sandboxed environment. This means the mini-apps are safe to run within the browser, and they can be saved, shared, or updated as needed. Google has indicated that these mini-apps will be stored in a user's account, accessible from any device. This feature could disrupt traditional app stores and software distribution models by enabling on-demand generation of utility software directly from a search query.
For example, a real estate agent could ask for a comparative market analysis tool that pulls local listing data and recent sales, and Search would produce an interactive dashboard. A student could generate a physics simulation to understand gravitational waves. The possibilities are vast, and Google is betting that this will make Search indispensable for both casual and professional users.
The AI race and the shift away from the open web
The aggressive rollout comes as Google aims to protect its territory against aggressive moves from rivals like Anthropic and OpenAI. While Google’s Gemini app boasts 900 million monthly active users, data from finance platform Ramp revealed that Google held just a 4.5% share of paid AI business subscriptions in the US this past April, lagging far behind Anthropic (34.4%) and OpenAI (32.3%). This disparity highlights the challenge Google faces in monetizing its AI capabilities and converting free users into paying customers.
The massive pivot toward self-contained AI answers has triggered concerns about the broader internet economy. Financial analyst Richard Kramer of Arete Research noted to The New York Times that the strategy helps Google maximize its lucrative ad platform but changes the nature of internet traffic. 'The open web is on its way out,' Kramer said, per the Times. 'With A.I., Google is reducing everyone to raw data providers.' This sentiment echoes fears among publishers and content creators that AI-generated summaries will reduce click-through rates and ad revenue for websites that depend on Google traffic.
Nevertheless, Google executives emphasized that the AI features are driving engagement to historic heights. During its first-quarter earnings call, Google reported that user queries reached an all-time high, pushing Google Search and other advertising revenues up 19% year-over-year to $60.4 billion. The company argues that AI enhances the search experience and encourages more interactions, which ultimately benefits advertisers and content providers through increased overall usage. However, critics point out that if users get answers directly from the search results page, they have less incentive to visit original sources, potentially starving the very content ecosystem that Google relies on.
The new intelligent search box is rolling out globally this week, while the information agents and custom application-building tools are scheduled to launch this summer, initially to Google AI Pro and Ultra subscribers. This staggered rollout mirrors Google's pattern of testing premium features before wider release. The company will face intense scrutiny from regulators and industry watchdogs, especially as antitrust proceedings continue regarding Google's dominance in search and digital advertising.
Looking ahead, the integration of AI agents and on-demand code generation could redefine productivity and information access. Users may soon rely on Google not just for answers but for active task execution and software creation. The shift also raises questions about privacy, as agents will have continuous access to user data and browsing behavior. Google has stated that all agent activities will be transparent and manageable through a new dashboard, and that no personal data will be shared without explicit consent. Yet, as with all AI systems, the potential for misuse or unintended consequences remains.
The AI race in search is far from over. Microsoft is rumored to be integrating more agent features into Bing, and Apple is reportedly developing its own large language model for Siri. The next few years will likely see an explosion of AI-powered tools that blur the boundaries between search, personal assistant, and software development environment. Google’s latest overhaul positions it at the forefront, but the competition is fierce and the implications for the open web are profound. As AI models become more capable and integrated, the very concept of a 'search engine' may evolve into something unrecognizable from its 1990s origins.
Source: TechRepublic News