Google is again attempting to change search with generative AI.
The cloud provider on March 5 launched Gemini 2.0 for AI Overviews in the U.S. Google first launched Gemini 2.0 last December as a family of multimodal language models.
Google also introduced an experimental AI Mode, a feature that expands AI Overviews with more advanced reasoning, thinking and multimodal capabilities.
AI Mode uses a custom version of Gemini 2.0 and is helpful for questions that need exploration, comparisons and reasoning, according to Google. Users can ask complex, multipart questions and follow-ups in AI Mode.
Changing search
The new feature is the latest in Google’s string of efforts to revamp its search model in the age of generative AI.
The cloud provider started this effort in 2023 with Search Generative Experience, a search and interface capability that integrated generative AI-powered results into Google Search.
In May 2024, SGE became AI Overviews, a feature that produces summaries of search results. In the year since Google introduced AI Overviews, the cloud provider has refined it in various ways, such as including links in the summary to take users to the point in an article that addresses their questions.
The latest iteration of AI search with AI Mode is a legitimate advance, said Mark Beccue, an analyst at Enterprise Strategy Group, now part of Omdia.
“This is the best progression we’ve made on advanced AI search in concept,” Beccue said, adding that he has not had a chance to use it yet, but has requested access. Google is currently inviting Google One AI Premium subscribers to be the first to experiment with AI Mode.
AI Mode is a logical step toward the next stage of generative AI search, Beccue said.
“It’s a vision we’ve been waiting for,” he said. “Whether it will work, we don’t know yet.”
The enterprise angle and challenges
While AI Mode and AI Overviews will be useful for consumers, they can also help enterprises, said Arun Chandrasekaran, an analyst at Gartner.
“Both AI Mode and AI Overviews can enable Google to serve more succinct, actionable search results while hopefully providing capabilities for complex search questions,” Chandrasekaran said. “This could be valuable in the enterprise for search in complex domains like R&D, engineering and legal.”
However, while improving AI search is Google’s main goal with AI Overviews and AI Mode, the cloud provider is also defending its leadership position in search. In the past year, competitors in AI search have included OpenAI, with ChatGPT search; Perplexity AI; and Meta, which is reportedly exploring generative AI search.
“These announcements are aimed at establishing Google as an AI leader that is receptive to market demands,” Chandrasekaran said.
But the cloud provider must serve two constituencies in the search market: users and advertisers.
“They had to think of a way to go into their model, which is search, paid search, and [figure out] how do you support paid search so that people, the customers, get what they want, and the advertisers get what they want,” Beccue said. “They’re looking for the best way to do both.”
While Google has been challenged by Perplexity and ChatGPT search, it has retained its dominance in search, holding nearly 90% of market share. But it has been looking to make its search model profitable, Beccue continued.
The challenge Google faces shows the daunting task of “effecting user behavior change and exploring new revenue models with AI-powered search,” Chandrasekaran said.
“Google is simultaneously grappling with the innovator’s dilemma and pivoting its business model and user experience,” he said.
With Google taking this next step in AI search, questions remain about how others such as Yahoo — the longtime email provider that has maintained a presence in search and another website that offers search capability — and Microsoft will compete.
Esther Shittu is an Informa TechTarget news writer and podcast host covering artificial intelligence software and systems.
