In a recent interview, Google CEO Sundar Pichai explained how search is evolving in response to advances in AI. The discussion focused on a simple question: if AI can act, plan and execute, what role will research play in the future?
Information requests can become an AI agent search
The interviewer asked whether search remains a product or becomes something else as AI systems begin to handle tasks instead of returning results.
They asked:
“What do you see as the future of research? Is it a delivery mechanism? Is it a future product? Is it one of N ways people will interact with the world?”
If Pichai had been interviewed by members of the publishing and SEO community, his response may have received pushback. He responded that search is not being replaced, but continues to develop as new features are introduced and user expectations evolve.
He said:
“I feel like in research, with every job change, you can do more.
And we need to absorb these new capabilities and continue to evolve the product frontier.
If it’s mobile, the product has evolved quite quickly, you come out of a New York subway, you look for web pages, you want to go somewhere, how do you find it? So you’re constantly evolving, people’s expectations are changing, and you’re moving forward.
If I fast forward, a lot of what are just information search queries will be agent searches. You will be performing tasks, you have many threads running.
In the first example of someone exiting a New York subway, yes, someone can search for a web page, but will Google show the user a web page or treat it as data by summarizing it?
The second example completely removes the user from the search and inserts agents in the middle. This scenario implicitly treats web pages as data.
Will research exist in ten years?
Pichai was asked what the future of research would be in ten years. His answer suggests that the future of search will involve many information-seeking queries being treated as tasks performed by agentic AI systems. Additionally, search will become more like an orchestration layer between the user and AI agents.
The exact question he was asked was:
“Will research exist in ten years?
The Google CEO responded:
“It continues to evolve. Search would be an agent manager, right, where you do a lot of things.
I think in a way, I’m using anti-gravity today, and you have a group of agents that are doing things.
And I can see research doing versions of these things, and you realize a bunch of things.
At this point, the interviewer attempted to get Pichai to return to the question of the true research paradigm, whether it would exist in ten years. Pichai declined to specifically state whether the research paradigm would still exist.
He continues his response:
“Today, in AI search mode, people are doing deep searches. So that doesn’t quite fit the definition of what you’re saying. But those kinds of people have adapted to that.
So I think people will be doing long-running tasks, which might be asynchronous.
What he described is a version of search that handles actions across multiple steps, in which multiple processes can run simultaneously instead of returning a single ranked set of results. And yet it’s strangely abstract because it talks about queries but fails to mention websites or web pages in that specific context.
What is happening? His next response highlights this more clearly.
Who is the flea and who is the dog?
The interviewer picked up on Pichai’s mention of adaptation, made an analogy with evolution, then asked:
“It’s almost like this old version or this paradigm ends up disappearing? And what was search becomes an agent and your future interface is an agent, and the search field in ten years or n years is no longer the…”
Pichai interrupted the interviewer to say that it is no longer possible to predict five or ten years because patterns change, what people do changes quickly, and given this pace, the only thing to do is accept it.
He explained:
“The form factor of devices is going to change. The I/O is going to change dramatically. And so… I think you can paralyze yourself by thinking ten years ahead. But we’re fortunate to be at a time where you can think a year ahead, and the curve is so steep. It’s exciting to do that year, right?”
Whereas in the past you might have to sit down and look at five years, as opposed to models that will be drastically different in a year.
…I think it will evolve, but it’s a moment of expansion. I think what a lot of people underestimate in these moments is that it seems so far from being a zero-sum game, right? The value of what people are going to be able to do also follows a crazy curve, right?
I think the more you look at it as a zero-sum game, the harder it seems. It can become a zero-sum game if you innovate or the product doesn’t evolve.
But as long as you’re on the cutting edge of these things, and we’re doing both research and Gemini, so they will overlap in some way. They’re going to differ profoundly in some ways, aren’t they? And so I think it’s good to have both and embrace them.
What the Google CEO is doing is rejecting the possibility of becoming obsolete by deliberately focusing on competitive agility and viewing uncertainty as a strategic advantage.
This might work for Google, but what about websites?
I think businesses also need to embrace competitive agility and move away from the flea-on-the-dog mental attitude. And yet, online businesses, publishers and the SEO community are not fleas, because it is Google itself that feeds on the web’s content.
What about websites?
The interview lasted more than an hour and at no point did Pichai mention any websites. He mentioned web pages twice, once as something to understand with technology and once in the example of a person exiting a subway looking for a web page. In both of these cases, the context was not that of a Google search searching for or retrieving a web page in response to a query.
Given that Google Search is used by billions of people every day, it’s a bit strange that the websites aren’t mentioned at all by the CEO of the world’s most successful search engine.
Watch the video starting around eight minutes in:
Featured Image/Screenshot from Video
