Google said its new AI search features can use advanced reasoning to answer search queries with structured, actionable responses that curate text and images from multiple sources. The approach would mark a notable shift away from the familiar search engine results pages the company is known for.
More people, smarter answers
AI Overviews, a popular feature in Google Search, is receiving a major performance upgrade with Gemini 2.0, Google’s advanced AI model. This upgrade enables the feature to handle more complex queries, including coding problems, advanced mathematics, and multimodal questions, which involve various types of data like text and images. Google says responses will now be faster, more accurate, and richer in detail.
As part of this expansion, Google is removing sign-in restrictions for teenagers, making AI Overviews accessible without needing to sign in. The company hopes this will help younger users with educational queries and research.
A new experimental AI Mode
For users who want even more AI assistance, Google is testing a new AI Mode in its Labs section. This experimental feature takes AI Overviews a step further by offering advanced reasoning, deeper analysis, and multimodal responses to handle more nuanced, multi-step questions.
For example, users can ask, “What’s the difference in sleep tracking features between a smart ring, smartwatch, and tracking mat?” Instead of fragmented results, AI mode provides a detailed, structured breakdown with follow-up options and curated links for further exploration.
AI Mode uses a technique called “query fan-out,” which means it conducts multiple related searches at once across different data sources and combines the results into a single, comprehensive answer. This approach is particularly useful for complex tasks like product comparisons, conceptual explorations, and in-depth research.
Google emphasizes that AI Mode isn’t just about generating answers — it’s about connecting users to the best information available. The feature integrates Google’s Knowledge Graph, real-time data, and shopping information for billions of products, ensuring users receive accurate responses with direct links to reputable sources.
Testing the waters
Google has been testing AI Mode internally and with select users, gathering feedback on speed, accuracy, and fresh content delivery; now, it’s expanding testing through an opt-in experience in Labs, allowing Google One AI Premium subscribers to be among the first to try it.
The company acknowledges that the system is still in development and won’t always get things right, especially when handling subjective topics. To address this, Google is fine-tuning its models to ensure answers remain objective and grounded in factual data.
In the coming months, Google plans to introduce more enhancements, including visual responses with images and videos, richer formatting, and improved ways to connect users with helpful web content.