Google has launched AI Overviews powered by Gemini 2.0 in the US, enhancing responses to complex queries like coding and advanced math. While claiming over a billion users, the service is mandatory for those using Google search, now affecting teens and untethered users. Additionally, Google is introducing AI Mode as an experimental feature in Search Labs, allowing users to opt in for deeper exploration and nuanced question handling. It integrates advanced capabilities with Google’s information systems but may unintentionally convey opinions. Future updates will enhance visual content and improve response quality, with Google One AI Premium subscribers getting early access.
Google has officially announced that AI Overviews in the US are now enhanced by Gemini 2.0. This update aims to tackle more complex inquiries related to coding, advanced mathematics, and multimodal queries, with Google assuring users of quicker and higher-quality responses.
Interestingly, Google claims that over a billion people are “using” AI Overviews, which feels misleading since it’s not an optional feature; it gets pushed to users whenever Google decides. This means those so-called “users” lack any real choice. In a twist, the update expands access to AI Overviews for teenagers and individuals not signed into their Google accounts. If you were trying to bypass AI Overviews, that workaround is now ineffective. Welcome to the new era.
Furthermore, Google appears to believe there’s still room for more AI on your search pages. As such, it’s rolling out AI Mode as an experimental feature in search (for the time being). This is something you can actually opt into via Search Labs.
This mode amplifies the capabilities of AI Overviews with “enhanced reasoning, critical thinking, and multimodal features,” essentially hitting all the latest AI buzzword trends. The tab for this new mode positions itself at the far left, preceding even the “All” tab. Within this mode, users can pose questions and engage in follow-up inquiries along with “helpful web links” (at least for now).
AI Mode employs a tailored version of Gemini 2.0 and is “especially beneficial for inquiries requiring deeper exploration, comparisons, and reasoning,” according to Google, which seems to have a vague grasp on what “reasoning” or “thinking” entails (though that’s not a solely Google issue). The highlight here is that “you can pose nuanced questions that might have previously necessitated several searches,” yet you’ll still need to verify all the responses with multiple searches for reliability.
This experience consolidates “advanced model capabilities” with Google’s “top-tier information systems,” whatever that implies. Users can access high-quality web content while also leveraging “up-to-date, real-time sources like the Knowledge Graph.” There’s even shopping information for billions of products because why not?
And here’s the typical fine print: “while we strive for AI responses in Search to present information neutrally based on online resources, some responses may inadvertently seem to adopt a persona or express a specific opinion.” But don’t fret; all of this will be tackled in the forthcoming testing phase. Also on the way: more visually enriched responses featuring images and videos, enhanced formatting, and “new methods to find useful web content,” whatever that entails.
Google will begin inviting Google One AI Premium subscribers to be among the first to experience AI Mode.
Source