What is search, anyway? At Google I/O this week, the company made that question even harder to answer, as Search absorbed more of its Gemini AI capabilities and moved further beyond the familiar list of blue links.

For decades, users directed search engines to fetch a list of pages where answers could be found. Now, search engines are acting more like butlers, anticipating what users want before they want it, based on what they know about them.

It’s not too much of a stretch to anticipate that “search” and “AI” are blurring, and will likely simply merge at some point. As announced at I/O, the Google Search tool is soaking up even more AI capabilities, with an expanded search box plus personal agents, while Gemini itself is taking on more tasks, such as delivering a daily brief, that were normally associated with a personal aide or attaché.

Today, what might have been called “notifications” in past eras are now the domain of agents. Google wants users to ask them to keep their eyes open for nearly anything—low plane fares, news about a favorite artist, updates from an apartment complex, and so on. Two key additions are search agents and a personal agent called Spark.

It’s hard to separate one from the other, just like it’s a challenge to distinguish Gemini within Google’s Workspace apps from the Gemini app itself. They are simply blending together.

The rise of agents continues

Spark is Google’s new “24/7, personal agent” that works on behalf of users. Right now, it is apparently somewhat basic, allowing users to set recurring tasks or triggers, or teach it skills like checking an inbox for updates from a child’s school. Over time, however, Google has a roadmap of features planned for Spark, just like for any of its other properties.

Of more use, perhaps, is what Google calls a “daily brief.” If that sounds familiar, it might: Microsoft built a daily summary of upcoming events into Windows 10’s Cortana, and later tried to move this feature into the mobile Outlook app. Which one would have come out on top is unclear, but the effort deserves applause.

“It goes far beyond a simple summary,” according to a blog post by Josh Woodward, vice president of Google Labs and the Gemini app. “Daily Brief actively organizes and prioritizes based on your specific goals, even suggesting immediate next steps.”

Of course, there’s no telling whether the daily brief will prove effective. Not surprisingly, it benefits from connections to Gmail, calendar, and other connected Google apps. It also requires a subscription, though it is available to the AI Plus tier as well as the more rarified Pro and Ultra subscription models.

Search and Gemini: Are they on a collision course?

As one might imagine, AI Mode—Google’s controversial revamp of its search function—has drawn mixed reactions. Like virtually everything else in Google’s ecosystem, Google Search now includes “personal intelligence,” mining user life for additional context. Google now says that AI Mode, which grudgingly links back to the original source of its knowledge, has landed one billion users.

Google is expanding the search box, literally, at least on its mobile implementations. This will allow for longer, more involved queries where users can see the entirety of the prompt, add files, and so on. The upshot is that Google does not want users to search for “best laptop”; it would rather they input something like “the best laptop like the one my cousin Mike had last summer at the house in Maine, but under $1,500,” with everything from text, images, video, to even Chrome tabs as potential inputs.

Google Search is encouraging conversations with its search engine, rather than one-off queries. At this point, the line between a “search” and a “prompt” blurs even further, especially when Search and AI Mode allow follow-up conversations. AI Mode, heavy on the “AI,” is already live where implemented.

We can see the impact of AI Mode on the business, and it isn’t good. But we cannot be as negative about Google’s agentic search capabilities.

Generally speaking, the industry has come up with numerous ways to facilitate ongoing searches. Steam and Amazon offer the ability to “wishlist” a particular item, tracking its pricing and notifying when it is on sale. Microsoft still implements “Collections” of stored tabs, where users can research and store an ongoing project, like a summer vacation. (Like Google, Microsoft is pushing users to adopt Copilot to take over manual tab storage and replace it with an AI summary.)

Agentic search used to be called “notifications,” where users could tell Google to monitor a topic and it would track it for them. Now Google Search is adding “search agents,” which will essentially monitor an existing question and provide answers.

“With information agents, you can stay updated on whatever matters most to you,” according to a blog post by Liz Reid, the head of search for Google. “Your agent will intelligently look across everything on the web, including blogs, news sites, social posts, and the freshest data, such as real-time information on finance, shopping, and sports, to monitor for changes related to your specific question.”

Beginning this summer, users will also be able to allow Google to reserve restaurants and other venues—and even pay. The latter has been a capability where AI has feared to tread, but Google is setting out to make it happen.

Google is also using its own version of Claude Code, called Antigravity, to build small “apps” right within search itself. Google is not really building an app that will allow users to do something. Instead, it is using Antigravity to create small visual explanations of how a specific task could be completed or how a concept actually plays out in the real world, such as a black hole’s effect on time and space or how a Roman aqueduct may have been constructed.

We do not have any vacations planned for the near future, but agentic search could absolutely be used to answer questions like “who is the current leader in the California governor’s race polling?” or “how much money has OpenAI raised in 2026?”

Like it or not, Google is one of the architects of the modern search experience, and how we look for and acquire information. Anecdotally, Google still has 90 percent of the world’s search traffic, according to StatCounter. The issue, of course, is how the problem is defined—how many people are simply “searching” via ChatGPT or Claude? Ongoing agent-based searches and conversational follow-ups will keep users within Google’s fold, where its management is desperate that they remain.

This transformation is not happening overnight. The integration of AI into search has been building for years, with Google first introducing neural matching in 2015, BERT in 2019, and MUM in 2021. Each step made the search engine better at understanding natural language and context. Now, with Gemini, the company is aiming for a seamless blend of information retrieval and task execution. The ultimate goal is to make Google the single place where users not only find answers but also take action—booking trips, making payments, tracking projects, and managing daily life. The line between search and AI assistant has become so thin that users may soon forget it ever existed.

Source: PCWorld News