Google releases Gemini Generative AI prompt guides

Google has released two new Gemini generative AI prompt guides for use on Google Cloud and Google Workspace. Workspace is the company’s collaboration and productivity software suite (formerly GSuite), while Gemini can be used as an assistant within the Google Cloud Platform console. The guides help users write effective prompts to increase productivity, efficiency, and quality of work – and help automate routine tasks.

The guides are part of Google’s new effort to increase generative AI skills for users of its Workspace and Google Cloud platforms. They are in addition to a new blog series, “Beyond the Prompt,” offering continuous productivity tips for Gemini.

Why use Generative AI prompts?

If you’re just starting to experiment with generative AI, prompts are used in everything from chatbots like OpenAI’s ChatGPT to text-to-video tools like OpenAI Sora and text-to-image tools like Adobe Firefly.

But what exactly is a prompt? Think of it as a conversation starter with your AI-powered assistant. Like any good conversation, you might have follow-up questions, requests, and thoughts.

How a user prompts a generative AI tool can make all the difference in the output. Effective prompts vary based on the model used, but some general best practices are helpful.

Download the free Google Gemini prompt guides

The Google Gemini generative AI prompt guides are available on Google’s website:

• Google Gemini for Workspace Prompt Guide (for Google Workspace)
• Google Gemini Prompting Guide (for Gemini used across Google Cloud)

Additional helpful resources for Google Gemini can be found across their documentation portal. Here are a few helpful links:

• Writing better prompts for Gemini (for Gemini used across Google Cloud)
• How Gemini works
• Google Gemini For Developers: Gemini Cookbook (example prompts, code, and integrations)
• Use Gemini for AI assistance and development
• Setup Gemini Code Assist for a project (using Gemini to help inspect and write code, and offer suggestions for code optimization, and build integrations to other services)
• Setup Gemini in BigQuery (for data warehouse and data workflows like generating SQL and Python code, get recommendations on partitioning and clustering, etc.)
• Setup Gemini in Databases (for database-specific workloads and tasks like generating SQL queries, fine-tuning database performance, etc.)

If you’d like to experiment more with prompting techniques, we recommend using Google AI Studio or NotebookLM. Google AI Studio will allow you to work with different controls and models. At the same time, NotebookLM is excellent for attaching multimodal inputs like PDFs, documents, and images, and it simulates retrieval augmented generation for grounding responses.

Finally, bookmark the Google Gemini community forum for additional resources and support from fellow developers.