Using AI technology, Google is changing its search field for the first time in 25 years

For 25 years, Google’s iconic search box was a long, thin strip where people typed in keywords like “World Cup.”

But over the past three years, artificial intelligence has allowed humans to ask longer and more complex questions like “Who are the top 24 teams in the World Cup and what chance does the United States have to advance?”

On Tuesday, Google said the AI ​​shift inspired it to redesign the dimensions of its search box for the first time since 2001. The box is getting bigger and more interactive, so people can ask even longer questions and upload photos and videos to queries.

Additionally, people can ask follow-up questions with the chatbot on the main Google search page. The company will also offer digital assistants, known as agents, to automate searches so that someone who might be looking for an apartment can be notified of a new listing without having to open a real estate site like Zillow.

The search function will be powered by the new Gemini 3.5 Flash artificial intelligence model. Google said the model improved at creating software code and performing autonomous tasks, worked faster and was cheaper to run than comparable models.

Sundar Pichai, Google’s CEO, said that Gemini’s speed and affordability have made it possible for them to be widely available – which will ultimately benefit Google.

“When people use our AI-based features to search, they use more searches,” Mr. Pichai said in an interview Tuesday before Google’s annual developer conference. The changes to search and Gemini 3.5 anchored a nearly two-hour demonstration of the company’s AI-powered products, including a new video tool, an online shopping cart and a system for autonomously reading and drafting emails.

Google has increasingly narrowed its AI lead against rivals such as OpenAI and has begun to struggle for leadership. After OpenAI released ChatGPT in 2022, Google faced concerns that the AI ​​startup would erode its search dominance. Concerns deepened after one of its early AI products recommended that people use glue to make pizza.

Last year, however, Google solidified its position as an AI heavyweight. In addition to its Gemini models, it made AI chips and poured hundreds of billions of dollars into data centers for its cloud computing business. Its coding and research app Gemini now has 900 million active users — about the same number as ChatGPT.

Google is using artificial intelligence to tap into other pockets of the digital economy. Web summaries generate more searches, while longer queries provide better insight into users, and new shopping features make it easier to connect customers with retailers.

Richard Kramer, a financial analyst at Arete Research, said the changes helped Google make more money from advertising. Last year, clicks on Google ads increased by 6 percent and were charged 7 percent more for each click. The company’s annual profit has more than doubled to $132 billion by 2022.

“The open web is on its way out,” Mr. Kramer said, referring to the way Internet traffic now often begins and ends with a visit to Google, rather than a visit to other sites. “With artificial intelligence, Google reduces everyone to providers of raw data.”

Google’s AI transformation stands out especially in search. In 2024, the company stopped filling some queries with a list of websites and instead provided automatically generated answers called AI Overviews. Last year, it added a search tab called AI Mode, where people can ask multiple questions on the same topic as with the chatbot.

Google said these features were combined. In searches that provide AI insights, people can ask follow-up questions in the AI ​​mode, which Mr. Pichai called “reveal.”

Google is also bringing one of AI’s biggest breakthroughs — software coding — to search. When people research complex topics like astrophysics, Gemini can create interactive graphics and simulations behind the scenes that provide a deeper answer than the previous list of websites.

This feature builds on the latest products from Anthropic and OpenAI, which have created tools that automatically complete code and create agents to automate email, research, and office work.

Google said it is rolling out an alternative to agents using Claude Code and Anthropic’s OpenAI Codex. Called Gemini Spark, the service is built into Gmail, Docs and other Google products, where it can turn meeting notes scattered across emails and chats into a single document. It can also read and draft emails.

The company also unveiled an update to its AI coding platform. It’s called Antigravity 2.0, and Google said it will have access to Gemini 3.5 Flash and will be able to bring huge cost savings to companies that process huge chunks of AI data every day to write code.

While many AI labs have pushed their models to accomplish more complex tasks, Google has focused on integrating Gemini deeper into its already popular services like shopping and YouTube. Shoppers can now create a cart in search or on YouTube while browsing products, instead of having to save items to purchase on a merchant’s website.

Google’s AI-powered shopping cart will also recommend discounts when products go on sale and warn people when they select items that might be incompatible with each other, such as choosing the wrong chips when building their own PC or the wrong filters when buying a coffee maker.

Google also plans to bring Gemini to photo editing. The company has created an editing tool called Gemini Omni that will let people, for example, change a vacation video in the Gemini app by telling the system to remove someone from the background. The company said editing tools will also be available in its Photos app in the future.

Omni also works as a video generation tool. Google said it can generate Hollywood-quality 10-second videos using prompts such as asking for “videos that explain a passage from a textbook” or “draw an imaginary character from a sketch.” Unlike Sora, OpenAI’s free video generator that has been discontinued, the Omni video generator will only be available to subscribers to one of Google’s AI services, which range from $8 to $250 per month.

(The New York Times has sued OpenAI and Microsoft, alleging copyright infringement of news content related to AI systems. Both companies have denied the suit’s claims.)

Koray Kavukcuoglu, chief technology officer of Google DeepMind, the company’s AI lab, said that Gemini’s involvement in Google products will help the company stay ahead of the competition with information about user needs.

“This feedback, the signal we’re getting, is the most important information flow we have,” he said.

Google also said it will bring Gemini to glasses with Samsung Electronics and glasses makers Warby Parker and Gentle Monster this fall. The glasses, which will work similarly to Meta’s Ray-Ban smart glasses, come with a camera, microphone and speakers so people can ask Gemini questions about their surroundings, such as the name of a landmark they’re staring at.

Dave Gilboa, co-CEO of Warby Parker, said he wore the glasses to ask Gemini for advice on installing a new car seat for his 3-year-old daughter, and also used them when she asked the question “Why?” about all sorts of things.

“These glasses were a massive unlock,” he said. “I can actually get her accurate information for her questions.