Google AI Mode will reinvent Search. I’m worried — and you should be, too

Table of Contents

What is AI Mode for Google Search?
A rich history of risks
This is not the Search evolution I seek

Google is moving forward with integrating more AI into internet search functionality. Remember AI Overviews, which provided summaries of website content at the top of the Google Search page? This feature, powered by the new Gemini 2.0 AI models, is now expanding to the US market without requiring a Google account sign-in, welcoming users of all age groups. However, Google is introducing a new AI Mode that transforms the traditional website link-based search experience into a conversational dialogue, akin to AI chatbots providing direct answers. While this offers convenience, the history of AI Overviews indicates potential errors.

What is AI Mode for Google Search?
The main concept behind AI Mode is to offer users information from indexed websites without the need to click on sources or navigate through webpages. Users can ask questions in natural language and receive detailed answers without conducting multiple searches. Google explains that AI Mode uses a “query fan-out” technique to simultaneously search across subtopics and data sources, presenting a comprehensive response. Despite positive internal test results, Google warns that AI Mode may not always provide accurate responses. In such cases, the traditional web search results will be displayed.

AI Overviews have a history of inaccuracies, highlighting challenges in the AI landscape. Missteps like AI hallucinations, where AI tools fabricate information as factual, are prevalent. For instance, an AI Overview erroneously stated a right-side driving rule in India. The condensed information provided by AI Overviews can mislead users without prompting fact-checking. These inaccuracies may pose risks, especially in health-related queries. AI misinformation, like suggesting harmful practices, could have severe consequences.

This is not the Search evolution I seek
Despite Google’s claims of AI model advancements, inaccuracies persist. AI Overviews have provided misleading information, such as incorrect Grammy award counts and wrong years. The AI Mode’s response to simple queries like the current year can be flawed, presenting irrelevant details. While AI tools like Gemini Deep Research are beneficial, implementing error-prone AI into Google Search raises concerns about information reliability. DailyTech has contacted Google for a response.