Google's AI Overviews: A Sneak Peek at the Future of Search

Over the past year, one of the biggest questions surrounding Google has been about its main product and primary revenue source: Will AI chatbots replace search engines? In May, Google provided some answers. The company announced that in the next era of search, AI will handle the work, so users don’t have to. A video revealed that AI Overviews, the new term for AI-generated answers, would soon appear at the top of search results. This marks a gradual move toward a future where the internet doesn't just offer links and hints, but direct answers.

Any update to Google’s search engine is a significant event. The search box is a primary gateway for people to access the internet on their computers and phones. This recent tweak has been portrayed as a major milestone in the media, given Google's enormous, often controversial influence as both a distributor and monetizer of online attention, which could be on the brink of transformation.

After almost a year of testing, however, Google’s AI-search experiment has seemed to me less like a revolutionary change and more like another questionable addition to an already cluttered results page. These days, I glance at the AI-generated responses just long enough to see that they’re occasionally glaringly incorrect. Perhaps it will improve, as Google promises; perhaps accuracy doesn’t matter if users find it appealing regardless. Either way, the debate over whether Google is about to overhaul the entire web economy, and whether these AI summaries will deliver a fatal blow to publishers and other platforms reliant on Google, won't remain unresolved for long.

Google's ambitions for AI in search are clear: to stay ahead of competitors like OpenAI and maintain its dominance. However, search is just one aspect of what the company showcased at its annual developer conference, Google I/O, in May. The enhancements to search also signaled Google's deep commitment to AI, betting on its potential to redefine privacy norms, often to the advantage of corporations like itself.

At the event, Google introduced or hinted at various new tools for generating images, audio, and video. They announced a new voice assistant capable of answering questions based on what’s displayed on your device’s camera or screen. Upgrades were also revealed for assistants that can provide information about your documents, recent meetings, or email inbox. Additionally, Google is developing a program that can monitor phone calls in real time to detect language typically used in scams.

Some of these features are currently in the live-demo stage, while many others are still just ideas or marketing tactics. Google's message seems clear: "Whatever our competitors are doing with AI, we're doing it too, and we did it first."

A different narrative is emerging, though. Instead of seeing AI purely as a technology that Google is trying to figure out — are they its creator or victim, or both? — it’s becoming a continuation of a core company trait. A former CEO once described their policy as "Get right up to the creepy line and not cross it." Many AI tools explicitly state their benefits. In exchange, they seek fuller access to your digital life. The rush to deploy AI isn't just about innovation; it's about gaining more access and data, with an underlying assumption that users will comply.

This scenario isn’t new. In 2004, shortly after Gmail's launch, Google faced backlash for putting contextual ads in users’ inboxes, seen as a bold violation of privacy. Privacy advocates warned this was like "letting the genie out of the bottle." In hindsight, users accepted this trade-off, often without fully understanding it, setting a precedent for how the internet would operate.

By 2017, Google, which by then had a suite of data-dependent products, stopped scanning emails for ad targeting. This move towards privacy seemed somewhat absurd; Google's software was already on billions of phones, deeply integrated into users' lives.

Since then, shifts in privacy norms have been subtle. One day, a user notices their photo library has been automatically organized by faces. During a meeting, another user finds Zoom is transcribing their conversation. These small changes quietly reshape expectations around privacy.

AI assistants, which seem magical and are heavily marketed, offer tech companies a chance to push these boundaries further. These tools require access to data, often already granted by users. While it may not seem scandalous for a Google assistant to access Google Docs, it underscores how much control users give up.

Historically, Google has used unconvincing arguments to justify data collection, like showing more relevant ads. Users accepted or rejected these changes based on the software's benefits. AI assistants, however, make a more direct case for needing user data. They function better with more access, and although they're not fully here yet, this foreshadows potential privacy concerns when they arrive.

Google’s years of web data collection allowed it to generate search results. Similarly, AI assistants claim to operationalize this data for user benefits, turning personal data into a helpful chatbot. While this seems like a fair trade, it masks a deeper issue of choice illusion. Google acknowledges privacy concerns, like with its call-screening feature using on-device AI.

The idea that the AI boom threatens internet giants needs more skepticism. The tech industry's past, present, and future align logically. These companies thrived on collecting and monetizing user data. To unlock AI's full potential, they need more data. This isn't a conspiracy but an aspirational vision where traditional ownership notions are redefined. AI firms claim they need vast amounts of data to deliver on their promises. Google is making this personal: soon, it’ll help with everything, but it needs all your data in return.