Google is changing the way we search and, rather unsurprisingly, it’s now all about AI.
At its annual developer conference, Google I/O 2026, the company unveiled a huge overhaul of how search will work going forward. There are several big updates, which include a redesigned search bar, more Gemini integration, AI agents that can complete tasks for you and shopping tools that are designed to automate every step of the buying process, from price tracking to checkout.
On one level, the appeal here is obvious. If these new features work as efficiently as Google says they will, it’ll be a faster, more personalized way to search where AI helps handle the more boring and repetitive tasks so you don’t have to — that sounds genuinely useful to me.
But the more I looked through the announcements, the more it felt like Google wasn’t just redesigning search but the whole process of discovery itself. And I think that raises much bigger questions about critical thinking, the future of the open web and what happens when even more of our decisions are outsourced to AI.
The hidden cost of frictionless search
Google already offers both a dedicated AI Mode and AI Overviews, the AI-generated summaries that appear at the top of your search results. But what the company is announcing now feels much bigger.
The most significant shift might be the redesign of Google’s iconic search bar. It’ll look different and behave more like a chatbot, encouraging you to have a conversation with Gemini rather than type in traditional search queries. It’ll also make it easier to ask follow-up questions and trigger new AI-powered features.
One of the other new features allows AI agents to continuously monitor the web for whatever you’re looking for, whether that’s an apartment within your budget or tickets for an event. These agents can then alert you when the right result appears.
Another can help book services for you, like restaurant reservations or pet care appointments. Google is also experimenting with a universal shopping cart that pulls products from different retailers into one place
Looking at all of these new features as a whole, it’s clear what direction we’re heading in. Google no longer wants to simply help you search the web. It wants to filter information, compare results and increasingly act on your behalf. And for a lot of people, that will feel genuinely revolutionary.
Cognitive decline
I recently spent weeks checking apartment listings multiple times a day because email alerts never arrived fast enough. Having an AI agent constantly scanning listings for me probably would have saved me a lot of time and stress. Most people can probably think of some other repetitive digital task that’s been bugging them that they’d happily outsource immediately.
But the more I think about Google’s announcements, the more uneasy I feel about what else might disappear soon in the name of efficiency.
For years, searching for something online has involved a whole set of invisible mental processes. Like comparing sources, reading reviews, following strange rabbit holes, evaluating contradictions, deciding which information felt right and trustworthy and which didn’t. Sometimes even changing your mind halfway through.
AI search aims to compress all of that down into a single synthesized answer that requires very little from us. Like a lot of new AI tools, that certainly sounds convenient. But it may also fundamentally change our relationship with thinking itself.
We’ve covered some of the early concerns about the way over-reliance on AI is changing the way we think, work and feel. Earlier in the year, I spoke to journalist Ellen Scott about what she calls smoothout, the feeling of being disengaged, tired and demotivated at work because you’ve used AI too much.
And newer evidence is mounting. Researchers are starting to examine what happens when people offload too much cognitive effort onto AI systems, and it doesn’t look good.
One of the most widely discussed examples was a 2025 preprint study from the MIT Media Lab, which found that students using ChatGPT to write essays showed lower brain connectivity, weaker recall and less sense of ownership over their work compared to those using search engines or no tools at all.
Another study from Microsoft Research and Carnegie Mellon University found that people who trusted AI systems more heavily tended to engage in less critical thinking during workplace tasks.
The concern here isn’t that AI suddenly makes people unintelligent. But it’s that cognitive skills work a bit like muscles. So the less frequently you use them, the easier it becomes to rely on external systems instead while those skills slowly erode over time.
And Google’s vision of search seems to be built around reducing the amount of thinking users need to do themselves.
The problem with getting exactly what you want
But there’s another concern here too: what happens to the strange, messy and serendipitous nature of the open web?
Searching online can unearth hidden gems. I’m sure we all know what it feels like to go searching for one thing only to accidentally discover something else. Maybe you clicked through obscure forums, badly designed blogs, personal websites and niche communities. Sometimes the best discoveries came from not fully knowing what you were even looking for in the first place.
Even using my apartment example, the irony is I eventually chose somewhere that didn’t perfectly match the criteria I originally thought that I wanted. Which means that if I’d delegated the entire process to an AI agent optimizing for my exact inputs, I probably wouldn’t have even seen it.
That feels like the paradox of most highly efficient AI systems. They can become incredibly good at giving us exactly what we ask for while narrowing the chance for surprise, experimentation and genuine discovery.
And let’s not forget that sometimes “wasting time” researching things online is actually enjoyable. I know for me that it can feel immersive, absorbing and genuinely fun. That often gets lost in all of the conversations about AI and efficiency.
But the searching is part of the experience. Which is similar to the way people talk about AI-generated art or writing. Even if the end result looks impressive, many people would argue the creativity was never just about the finished product. The meaning came from the process itself. The experimenting, wandering, struggling, revising and discovering all kinds of things along the way.
AI search might eat the internet
There’s also a big contradiction underneath all of these new AI changes that I can’t stop thinking about, which is that Google’s AI search experience may end up reducing the need to visit websites directly. Publishers have already reported declining traffic due to AI summaries and chatbots that answer questions before users click through to the original source.
And yet the open web is also the very thing these AI systems depend on. Which means that if fewer people visit independent websites, fewer people have incentives to create them. And if fewer websites exist, the entire ecosystem supplying original reporting, reviews, ideas and information to AI systems could be threatened too.
That’s what makes Google’s AI announcements feel strange. Google appears to be cannibalizing part of the very web that made it powerful in the first place.
Maybe this is simply the next phase of the internet, where a smaller number of giant platforms control more and more of our interactions with information while what we knew of the web of the past becomes sort of obsolete.
Or maybe, what I secretly hope, is that people will eventually realize that convenience alone is not enough. Because some of the things being optimized away here were never meaningless inefficiencies. They were part of how we explored, discovered, evaluated and experienced information. So yes, AI-powered search will probably buy us back a lot of time. But at what cost?
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.
