Google Adds a New Gemini AI Feature That Will Let Users Fine-Tune Its Responses

Google has introduced a new feature for Gemini, its artificial intelligence (AI)-powered chatbot. The new feature provides users with granular control over the generated response to a prompt and allows them to make changes to a particular portion of the text. Users can regenerate the selected portion, edit its length or get it entirely removed. The feature is currently only available for the web app version and has not been extended to the app. Notably, Google recently announced multiple new AI features for Android devices.

Announcing the new feature in a post, Google said, “We’re launching a more precise way for you to tune Gemini’s responses. Starting in English in the Gemini web app, just select the portion of text you want to change, give Gemini some instruction, and get an output that’s closer to what you are looking for.” The feature is being added to offer more control to users’ creative process by letting them make changes to a small part of the response instead of regenerating it entirely.

Once a response for a prompt has been generated, users can select and highlight any part of the text. Highlighting it now will show a pencil icon in a circle. The Gemini logo also appears next to the pencil. Hovering the arrow over the icon brings the description “Modify selected text. Clicking on this icon shows four options — Regenerate, Shorter, Longer, and Remove.

The options work as the name suggests. Regenerate will change the portion of the text and it will be generated again. Shorter and Longer increases or decreases the length of the topic, and Remove deletes it entirely. Further, users can also add a prompt for that particular portion to modify it as well.

It is a handy feature. Those who have spent time on AI chatbots know that oftentimes, it generates a great response which is ruined by one sentence. So far, the solution to it was to regenerate the entire content or make modifications to the prompt, and repeat it ad infinitum till the desired result appears. Making granular changes via this feature can potentially save a lot of time for users.