Google’s Gemini Chatbot Learns to Analyze Search History

Google has trained its AI assistant Gemini to analyze and consider a user’s search history to provide more relevant responses.

The feature operates on the experimental “thinking” model Gemini 2.0 Flash Thinking. It references search results only when deemed “useful.” For instance, when choosing a restaurant, the chatbot will refer to recent food-related search queries to provide an answer.

This new solution is part of a broader push towards personalization by the corporation. In the future, Gemini will be connected to YouTube, Google Photos, and other company applications. This will enable the chatbot to “provide more personalized insights based on a broad understanding of user actions and preferences.”

Subscribers to Gemini and Gemini Advanced can enable search history analysis. The AI displays the personalized information it used.

The tool is available in more than 40 languages in “most” countries.

In March, Google launched an experimental AI Mode feature in its search engine, aimed at competing with Perplexity AI and OpenAI’s ChatGPT Search solutions.

Earlier, the corporation announced the addition of the ability to ask questions using video and display content on the screen in real-time to the Gemini AI assistant.