Google’s AI mode now allows users to ask complex questions about images
Google's AI Mode can now answer complex image-related questions.

Google recently introduced a new feature for its AI Mode that empowers users to ask multifaceted questions about images they upload or capture using their cameras. This development is part of their effort to integrate multimodal search capabilities into Google Search, allowing a deeper investigation into topics by addressing complex, multi-part questions.
The latest image-analyzing functionality leverages Google Lens' multimodal capabilities, as highlighted by Google in its Monday announcement. This integration enables AI Mode to comprehend the entire scene within an image, understanding intricate details like object relationships, materials, colors, shapes, and arrangements.
At the heart of this advancement is a technique called 'query fan-out.' Utilizing this method, AI Mode generates multiple questions pertaining to both the image itself and its constituent objects, thereby delivering more detailed information than a typical Google search might offer. A notable example mentioned is capturing a photo of a bookshelf and querying, "If I enjoyed these, what are some similar books that are highly rated?", prompting AI Mode to identify each book and suggest similar ones along with purchasing options.
Moreover, AI Mode facilitates follow-up questions, refining the search results based on user preferences – for example, asking for the shortest recommended book in a list of suggestions. Google's initiative includes extending AI Mode access to millions of users registered with Google Labs, beyond the previous availability restricted to Google One AI Premium subscribers.
Introduced last month, AI Mode positions itself as a competitive alternative to popular services like Perplexity and OpenAI's ChatGPT Search. Google has affirmed its commitment to enhancing user experience and broadening AI Mode's functionalities, indicating a sustained progression in integrating AI into everyday tasks.
Sources: Google Blog, TechCrunch, Verge