Google Chrome has unveiled a new AI writing tool, powered by Gemini, which is now available for users. This AI writing generator is an extension of the existing “Help me write” feature from Gmail and is now accessible across the web. It is driven by one of Google’s latest Gemini AI models and was initially announced in January. The tool is currently in its ‘experimental’ phase, requiring users to explicitly enable it.
Key Takeaway
Google Chrome has introduced an AI writing tool powered by Gemini, enabling users to create and modify content with contextual recommendations. While leveraging this feature, users should be mindful of the privacy considerations associated with the transmission of data to Google.
Enabling the AI Writing Tool
To enable the new writing feature, users can navigate to the Chrome settings menu and locate the ‘Experimental AI’ page. From there, they can easily activate the AI writer, along with Google’s new automatic tab organizer and Chrome theme manager. Presently, the AI writer is only available in English on Windows, Mac, and Linux. Once enabled, users can right-click on any text field and select ‘Help me write’ to create new content or rewrite existing text.
Contextual Recommendations
One notable feature of this tool is its ability to consider the context of the webpage being visited to provide relevant content suggestions. According to Google engineering director Adriana Porter Felt, the tool can extract key features from the product page a user is on, thereby enhancing the value of the generated content. For instance, if a user is crafting a review for a pair of running shoes, Chrome will incorporate essential product details to support the recommendation, making it more informative for potential shoppers.
Privacy Considerations
It is essential to be aware that the text, content, and URL of the page being used with the service will be transmitted to Google under its existing privacy policy. Google emphasizes that this information is utilized to enhance the feature, encompassing generative model research and machine learning technologies, including a review process involving human input.
