Google Chrome Tests AI “Skills” Powered by Gemini
▼ Summary
– Google is testing “Skills” for Gemini in Chrome, which will allow the AI to perform specific automated tasks based on user instructions.
– Gemini in Chrome currently acts as a helper, able to explain webpage content, summarize text, and compare information across open tabs.
– A future development will turn Gemini into an “agent” that can work tightly with Google apps like Calendar and Maps without switching tabs.
– The new “Skills” feature is accessible via a dedicated internal page (chrome://skills) where users can add tasks with custom names and instructions.
– This “Skills” functionality is currently in internal testing, and its public release timeline remains unclear.
Google is currently trialing a new feature called “Skills” for its Gemini AI within the Chrome browser, designed to automate complex tasks directly in your web browser. This development represents a significant step beyond the AI’s current role as a helpful assistant, moving it toward becoming an active agent that can execute multi-step actions on your behalf. The feature is being tested internally, with its public release timeline still under wraps.
For some time now, desktop users in the United States have had access to Gemini in Chrome. Currently, it functions as a smart companion. While browsing any webpage, you can activate Gemini to clarify confusing sections, provide concise summaries of lengthy articles, or even compare information you have open across several different tabs. Imagine planning a trip with separate tabs for flights, hotels, and local attractions; you could instruct Gemini to extract the crucial details from all three and synthesize them into a single, organized itinerary.
This new “Skills” capability, however, aims to take automation much further. As discovered within the browser’s code, a new internal page labeled `chrome://skills` is in development. This page will reportedly let users create and manage custom Skills by giving them a name and a set of specific instructions. These Skills would essentially be programmable commands that enable Gemini to perform targeted, repeatable actions within the Chrome environment.
The move aligns with Google’s previously stated goal of evolving Gemini into a more proactive “agent” over the coming months. The vision extends to a more integrated browsing experience where Gemini could help you locate previously visited webpages using simple, natural language questions. Furthermore, the AI is expected to work more seamlessly with other Google services like Calendar, YouTube, and Maps, potentially pulling relevant information into your workflow without the need to constantly switch between tabs or applications. This integration promises to streamline how users interact with both the web and their essential productivity tools.
(Source: Bleeping Computer)
