AI
Google DeepMind reimagines the mouse pointer as an AI-powered interface
The tech giant is developing context-aware pointer technology that lets users collaborate with AI seamlessly across their existing tools.
Google DeepMind Blog
Google DeepMind is rethinking one of computing's oldest tools: the mouse pointer. After more than 50 years of minimal change, the company is developing an AI-enabled version powered by Gemini that understands both what you're pointing at and why it matters to you.
The motivation is straightforward. Today's AI tools force users to copy content into a separate window to get help — breaking workflow and requiring lengthy, detailed prompts. DeepMind's vision flips this: intuitive AI that works wherever you are. Point to an image and ask for directions. Highlight a recipe and request doubled ingredients. No context-switching required.
Four guiding principles#
The team has outlined four core ideas shaping the new interface:
Maintain the flow — AI capabilities should work across all apps, not silo users into dedicated tools. The prototype pointer is available everywhere, whether in a PDF, spreadsheet, or email.
Show and tell — Instead of writing verbose prompts, users point at what they need and the system visually understands the exact word, paragraph, image section, or code block in focus.
Embrace "this" and "that" — Mirroring natural human conversation, users can make complex requests in shorthand ("Fix this," "Move that here") while relying on gesture and shared context.
Turn pixels into actionable entities — AI transforms visual elements—a handwritten note, a paused video frame, a photo—into interactive, structured data like to-do lists or booking links.
Rolling out now#
DeepMind is already integrating these principles into Chrome and a new Googlebook laptop experience. Users can now select items on a webpage and ask Gemini to compare them, or point to a room location to visualize furniture placement. A Magic Pointer feature is coming soon to Googlebook, with further experiments planned across Google Labs and Disco.
The work reflects a broader philosophy: building technology that adapts to human behavior, rather than forcing users to change how they work.

