Google Labs has announced Disco, a new experimental web discovery platform designed to explore how the web itself can better adapt to increasingly complex online tasks. As research, planning, and learning have expanded across more tools and information sources, Google says the traditional browsing model is struggling to keep up.

Disco is positioned as a new “discovery vehicle” from Google Labs, created to reimagine how people browse, build, and learn on the modern web. Through early experimentation and collaboration with early testers, the team aims to test new approaches that help users make sense of growing information complexity.

The first feature being tested inside Disco is GenTabs, built with Gemini 3, Google’s most advanced AI model to date. GenTabs aims to reduce the friction of modern web browsing—where research, planning, and learning often require juggling dozens of tabs at once—by transforming open browser tabs into interactive, AI-generated tools.

Unlike traditional AI assistants that rely solely on prompts, GenTabs proactively interprets what a user is trying to accomplish by analyzing their existing tabs and interactions.

From there, it generates custom, task-specific web applications that help move the work forward. Users can:

Because each generative element remains connected to the broader web, GenTabs preserves source attribution and allows users to trace information back to its origin.

According to Google Labs, early testers are already using GenTabs to create personalized tools for everyday needs, including:

These examples highlight GenTabs’ goal of turning static web content into adaptive, task-driven experiences.

GenTabs is built with Gemini 3, which Google Labs describes as its most intelligent AI model. The model enables GenTabs to understand browsing context, respond to natural-language input, and generate interactive web elements within Disco.

By integrating Gemini 3 directly into the browsing experience, Google Labs is testing how advanced AI models can support complex tasks without requiring users to leave the web or adopt new workflows.

Disco and GenTabs are currently available through a limited waitlist (see the link below), with initial access focused on macOS users. Google notes that the experiment is still early, and not all features will work seamlessly.

Feedback from this small group of testers will help determine which ideas are refined further—and which may eventually appear in broader Google products.

Q: What problem is Google Labs trying to solve with Disco and GenTabs?
A: Google Labs is addressing the growing complexity of web usage, where users often manage dozens of tabs to complete a single task. GenTabs aims to reduce cognitive overload by turning scattered information into focused, task-driven tools.

Q: How is GenTabs different from a traditional AI assistant or chatbot?
A: Instead of relying only on prompts, GenTabs understands a user’s intent by analyzing open tabs and chat history. It then generates interactive web applications directly tied to the user’s active browsing context.

Q: Do users need technical or coding skills to use GenTabs?
A: No. GenTabs is designed for non-technical users. Tools are created and refined using natural language, eliminating the need to write or understand code.

Q: How does GenTabs handle accuracy and source transparency?
A: Every generative element created by GenTabs remains connected to the web and links back to original sources, allowing users to verify information and maintain trust.

Q: Is Disco a finished product?
A: No. Disco is an early experiment from Google Labs. The company is starting with a small group of testers to gather feedback before deciding whether features like GenTabs evolve into larger Google offerings.

For everyday users, GenTabs points to a future where the web adapts to human intent instead of forcing humans to adapt to the web. Tasks that currently feel overwhelming—researching a topic, planning a trip, helping a child learn—could become faster, clearer, and less mentally exhausting.

For educators, parents, and knowledge workers, this approach suggests a shift from passive browsing to active sense-making, where information is organized around goals rather than links. That has implications for productivity, learning outcomes, and digital well-being.

More broadly, Google Labs’ decision to keep AI outputs tied to original sources matters. As concerns grow around AI-generated content eroding trust, GenTabs tests a model where AI enhances navigation and synthesis without severing the connection to the open web.

If experiments like Disco succeed, they could reshape how people experience the internet—moving from tab overload and fragmented attention toward AI-assisted clarity, while preserving transparency and user agency.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.