Google has unveiled an innovative AI experiment named “Disco,” powered by the Gemini technology, aimed at transforming open browser tabs into customized applications. This new tool introduces “GenTabs,” which proactively suggests interactive web apps tailored to your browsing activity, enabling users to build personalized apps using straightforward written prompts.
For example, when researching a topic, GenTabs can propose creating an app that visualizes key concepts, enhancing comprehension and engagement. Alternatively, users can leverage GenTabs for practical tasks, such as organizing a meal plan from various online recipes or planning a trip while exploring travel resources.
While some existing AI chatbots can perform similar tasks, GenTabs uniquely generates these customized applications in real-time using the Gemini 3 framework, drawing on the information available in your browser and your previous chats. Users can further refine these apps through natural language commands even after their initial creation. Google emphasizes that the generative elements produced will consistently link back to the original content sources.
As competition in AI technology intensifies, Google aims to infuse AI deeper into the browsing experience. Unlike standalone AI browsers such as Perplexity’s Comet or ChatGPT Atlas, Google has integrated its Gemini assistant directly into Chrome, which can respond to inquiries about the current web page. GenTabs not only focuses on individual page interactions but also addresses broader browsing tasks across multiple tabs, whether for research or leisure.
Currently, Disco will be available only to a select group of testers in Google Labs, who will provide feedback on the platform. Google indicates that promising ideas emerging from Disco could eventually enhance its larger suite of products. The introduction of GenTabs marks the first feature being trialed, with plans for additional functionalities in the future. Interested users can join a waitlist to access Disco, with the app rolling out on macOS initially.
