How AI queries are making Google search even smarter

This article summarizes an episode of Alex Kantrowitz’s Podcast featuring Google VP of search Elizabeth Reid and Google SVP of knowledge and information Nick Fox.

Photo credit: Shutterstock

Google’s generative AI appears to be expanding search usage, not cannibalizing it. According to Google SVP of knowledge and information Nick Fox and Google VP of search Elizabeth Reid, AI Overviews are prompting a 10% jump in queries on some topics because users now feel comfortable asking more complex questions.

Google’s strategy for this shift is a deliberate balance between rapid innovation and a cautious, trust-focused rollout. The company is reshaping search from a simple query tool into a multimodal assistant, altering its relationship with the web ecosystem it depends on.

Search is becoming a multimodal assistant

The interaction with search is evolving beyond simple keyword entry. Google is building a system that accommodates more natural and complex user inputs across different modalities. The goal is a fluid experience where users are not forced into a single method of inquiry.

• Conversational query: users are increasingly asking longer, more complex questions, moving away from fragmented keyword searches.
• Visual query: Google Lens is seeing significant adoption, allowing users to start a search with an image, such as a picture of a plant, and ask follow-up questions about it.
• Voice query: voice is becoming a preferred input for longer queries, especially on mobile devices, reducing the friction of typing.

Users should not have to adapt to the search engine.
Reid argues, “I don’t think use cases are all one or the other. I think we’ll see people interact fluidly… People will ask their questions in whatever way is natural. Like what is the way that you would like to ask that question? What’s the most effective way? What’s the most efficient way? Then we should make it easy for people to explore.”

Future interactions will be agentic and proactive

The evolution of search is leading to agentic systems that can act on a user’s behalf. This moves beyond providing information to accomplishing tasks, like tracking product prices or offering personalized suggestions. This shift requires models to master both reasoning and the use of external tools.

• Reasoning and tool use: models are improving at both reasoning through complex problems and using external tools, like querying real-time financial or sports databases, to ground their answers in factual data.
• Executing tasks: the system is moving toward proactive assistance. This includes features that can track a product’s price and notify the user when it goes on sale, directly executing a user’s intent.
• Content transformation: AI will increasingly remix content into new formats on demand. It can transform articles into short podcasts or create video clips, adapting information to the user’s immediate context and needs.

The new web ecosystem offers higher-intent traffic

Google’s strategy focuses on enhancing the value of the traffic it sends to the web. By using AI to handle simple informational queries, it directs users with more complex, higher-intent needs to publishers and advertisers. This process makes each click more valuable for the ecosystem.

• Expanded queries: generative AI is creating new search demand. Users are asking more complex, longer questions they previously would not have attempted, leading to an overall increase in search volume for certain topics.
• High-intent traffic: longer, more specific queries signal a user who is further along in their research or buying journey. This makes the resulting traffic more valuable to advertisers and publishers, as it is more likely to convert.
• Web-forward design: AI interfaces are being designed to be “web forward.” They prominently feature links and attribute information directly to sources, encouraging users to click through for deeper engagement.

Generative AI is a fuel for growth.
Fox states, “we see each of these reinvention moments as expansionary opportunities. We announced at Google IO that people are doing 10% more queries for the types of queries that trigger AI overviews. That’s a lot. That’s a meaningful change in user behavior. These are longer queries.”

Google’s strategy is trust over speed

Google is intentionally moderating its AI deployment speed. The company prioritizes user trust and factual accuracy over being first to market with every new capability. This approach accepts short-term competitive risks to preserve its long-term relationship with billions of users.

Maintaining user trust is the primary constraint on innovation.
As Fox says, “we care deeply about trust… we don’t want to make a mistake for our users. We don’t want to violate the trust of our users. So that’s the thing that to the extent that anything holds us back, it’s we don’t want to make a mistake for our users.”

A phased rollout de-risks radical change

Publisher success will depend on unique, expert content