Gemini, Google’s AI assistant, has obtained a side panel, which has been added to Gmail, Google Docs, Sheets, Slides, and Drive. The tech giant shared tips on how to get the most out of it, but only subscribers have access to the new feature.
Google recognizes that many people are still unfamiliar with generative AI, yet the tech giant aims to transform how we work day-to-day.
With the Gemini side panel announced a month ago, Google introduced multiple convenient uses for its AI assistant, such as summarizing documents, adding generated images, and drafting an email or generating other text.
However, the toolbar is only available for Gemini Business, Enterprise, and Education customers and Google One AI Premium subscribers. The plan costs $19.99 per month.
The Gemini assistant finds specific information in documents and emails. It summarizes, analyzes, and generates content by utilizing insights from these without switching tabs or applications.
“Gemini in Docs side panel can help you write and refine your content, summarize information, help you brainstorm, create content based on other files, and more,” Google said.
Gemini in the Slides side panel helps create new slides, generate custom images, summarize presentations, and more.
“We want to build a whole new level of connection between Workspace apps,” says Workspace Product Director Vishnu Sivaji. "Gemini is helping these apps become truly cross-product."
The scheduled release started on July 8th, and now Google is sharing five tips on how to interact better with the assistant.
First, the Gemini side panel provides product-specific suggestions, such as “refine,” “summarize” or “rephrase” the document.
“In Gmail, the side panel shows you ideas like “find attachments in my emails;” in Docs, you might see “rephrase part of this document;” or in Sheets, Gemini may suggest “teach me how to use conditional formatting,” Google writes.
Second, when interacting with Gemini, Google recommends focusing on the basics: what persona you want Gemini to be, what task it should do, add context, and describe the output format.
Like training a new team member, it’s better to use natural language, full sentences and specific requests, provide context, specify the needs, and even suggest a format.
“You wouldn't just say, ‘Sales report, please!’” Google explains.
“For example, you could write, “I’m a project manager” (persona) “and need to create a detailed project tracker” (task) “for a website redesign project” (context) “in a simple table with fields for dates, status, and tasks.” (format).
Third, users can use the @ symbol to give Gemini more information from other files from Docs, Sheets, and elsewhere in Workspace, without leaving the email or document.
For example, if a coworker sends you an email asking for some information, such as hotel details, to book a room, you can ask Gemini in the side panel to look it up from a specific document, like this: “What is the hotel name and sales manager email listed in @Company Offsite 2024?”
The information pulled by Gemini can be easily added to an email.
Fourth, experimenting and iterating help to get better results. Google suggests using different approaches, different keywords, and personas, adjusting the level of detail and specificity, and testing different prompt lengths if the result is not satisfactory.
“Based on the team’s research, the most successful prompts average around 21 words, yet people’s initial attempts are significantly shorter – usually fewer than nine words,” Google said.
For example, you may ask Gemini to act as a colleague but then compare the results to those asking Gemini to act as a teacher.
Fifth, Google thinks it’s important to share your findings with them, what’s working, and what could work better.
“The more input we get, the more helpful we can make our products.”
However, be cautious about sharing private data or documents with Google and review your privacy settings if needed.
Google states that user data stays in the Workplace and is not used to train or improve the underlying generative AI, Search, and other systems “without permission.” Its privacy policy also reveals that user interactions with Gemini are used to develop, train, fine-tune, and improve these models to better handle requests and update their classifiers and filters, including for safety, language understanding, and factuality.
Your email address will not be published. Required fields are markedmarked