Google Search is moving deeper into AI, and the company is describing 2025 as a major year for changing how people find information. CEO Sundar Pichai said during the company’s earnings call on Tuesday that Search is on a “journey” around AI, with AI overviews only serving as the opening step.
The broader direction is clear: Google wants Search to feel less like a page of links and more like an AI assistant. That means a product that can browse, examine web pages, return answers, support follow-up questions and, in some cases, handle research tasks that users previously did themselves.
Search is moving beyond simple answers
Pichai framed the shift as a major expansion of what people can ask Google Search to do. In his opening remarks, he said, “As AI continues to expand the universe of queries that people can ask, 2025 is going to be one of the biggest years for search innovation yet.”
That statement matters because it points to a change in Search’s role. The familiar model is built around people typing a query, scanning results and choosing which pages to open. The AI-heavy model is more active: Search can interpret more kinds of requests and produce a response that may reduce the number of steps a user takes.
The source article describes this as a long move away from a simple search system built around 10 blue links. Instead, Google is pushing Search toward a product that looks at the internet for the user and comes back with an answer.
This shift began publicly with AI overviews, which changed how Google delivers information to billions of Search users. Those summaries were controversial, but Google’s latest comments show that they were not the full plan. They were the start of a larger effort to pack Search with AI features from DeepMind.
DeepMind systems point to the next interface
One signal came when Pichai was asked about the future of AI and Search. He responded, “You can imagine the future with Project Astra,” referring to DeepMind’s multimodal AI system.
Project Astra can process live video from a camera or a computer screen and answer questions about what it sees in real time. In the context of Search, that suggests a future where the input is not limited to typed words. A user could ask about visual information on a screen or in front of a camera, and the AI system could respond based on what it sees.
Google also has plans for Project Astra outside Search. The company says it wants the multimodal AI system to power a pair of augmented reality smart glasses one day, with Google creating the operating system for them.
For Search, the important point is that Google is exploring AI systems that can handle richer context than a traditional query. Search could become an interface for asking questions about text, pages, screens and live visual information, rather than only matching keywords to indexed pages.
Research and web use could become automated
Pichai also pointed to Gemini Deep Research, an AI agent that takes several minutes to create long research reports. The feature matters because it automates work that people have traditionally done through Google Search: opening pages, collecting information and organizing it into a useful result.
As Pichai put it, “You are really dramatically expanding the types of use cases for which Search can work — things which don’t always get answered instantaneously, but can take some time to answer.” He added, “Those are all areas of exploration, and you will see us putting new experiences in front of users through the course of 2025.”
That is a different promise from instant search results. Some tasks may require time, synthesis and a multi-step process. Google appears to be looking at Search not just as a place to answer quick questions, but as a tool for completing more involved information work.
Project Mariner points even further in that direction. Pichai said Google has a “clear sense” of the Search experiences it could create with the AI agent, which can use the front end of websites on behalf of users. If Search can interact with websites directly, people may not need to navigate those sites in the same way themselves.
The implications are significant for two groups named in the source article:
- Websites that rely on Google traffic could face a Search experience where users get more done before clicking through.
- Businesses that buy ads on Google Search could see changes as the interface becomes more conversational and assistant-like.
ChatGPT pressure is shaping Google’s response
Google has been on this path for a few years, after being caught flat-footed by the release of OpenAI’s ChatGPT in 2022. Today, ChatGPT has matured into one of the internet’s most used products, with hundreds of millions of weekly users. The source article describes it as an existential threat to Google Search’s long-term business.
Google’s response has two tracks. It is building Gemini as a competitor AI chatbot, and it is also putting AI features directly into Search. That second track is especially important because Search is already the product people use when they want information from the web.
Pichai also said there is an “opportunity” in making it easier for people to interact with Search and ask follow-up questions. He said, “I think the [Search] product will evolve even more,” adding, “As you make it more easy for people to interact and ask follow-up questions, etc., I think we have an opportunity to drive further growth.”
That points toward a Search interface that behaves more like a chatbot, though Pichai was light on details. A user might ask a question, get an answer, refine the request and continue the exchange inside Search rather than starting over with a new query.
The first step showed the risks
Google’s AI push in Search has already had problems. When AI overviews rolled out to all of Google Search, the system produced inaccurate and strange AI hallucinations, including answers that told people to eat rocks and put glue on their pizza.
Google admitted at the time that AI overviews needed some work. That history is central to the next phase because the company is not stepping away from AI in Search despite the negative rollout.
Instead, the direction described by Pichai suggests that Google is preparing to put more AI experiences in front of users through 2025. The company’s Search product may still answer quick questions, but it is increasingly being shaped around AI agents, multimodal systems, longer research tasks and follow-up conversations.
The result could be a version of Google Search that is less about sending users to pages and more about doing web-based work with them or for them. That is the future Google is now openly sketching: Search as an AI assistant, not just a list of links.