Google Launches AI Research Tool Deep Research for Gemini Advanced Users

Google has announced that its experimental AI research tool, Deep Research, is now available to Gemini Advanced users via the Google Workspace platform.

Deep Research is a new feature unveiled last year designed to explore complex topics in depth and conduct detailed research, answering highly specific questions posed by users. The tool gathers and analyzes information from hundreds or even thousands of sources online, presenting its findings in a comprehensive and easy-to-understand report format.

The primary advantage of this tool is its ability to complete tasks in minutes that would typically take human researchers hours, making it suitable for various business and educational scenarios.

In the business sector, Deep Research aids in industry analysis, helping companies better understand emerging trends such as sustainable battery technology, thus informing their clean tech strategies. It can also be used for competitive analysis and customer research, enabling businesses to analyze local and global competitors or determine optimal locations for new small enterprises. Sales teams can use it to create detailed reports tailored to specific customers for company products and services.

In education, educators can leverage Deep Research for curriculum planning, classroom project development, presentations, and grant applications, generating reports on nearly any topic.

Currently, Deep Research is only accessible to paid Google Workspace users who have access to Gemini Advanced. Google Workspace is Google's productivity suite, including tools like Google Docs and Sheets. Gemini Advanced is the premium version of the Gemini AI application, supporting more tokens than the standard version and offering additional benefits such as increased storage space.

Gemini Advanced users need to switch to the "1.5 Pro with Deep Research" mode and input a simple prompt about the topic they wish to explore in-depth to start using Deep Research. The tool creates a multi-step research plan that users can modify as needed. Once ready, it immediately begins analyzing the most relevant information found through web searches, continuously refining its analysis based on what it learns. For instance, if researching new trends in clean batteries and discovering a promising new technology in the field, it will automatically dig deeper into as much information as possible about the new technology.

Upon completion, Deep Research generates a comprehensive report summarizing key insights learned. The report is presented in an easily digestible format and includes links to all important original sources, allowing users to explore different aspects further if desired. Users can also ask follow-up questions for deeper investigation.

Google claims that Deep Research can significantly save time, as thorough research is widely recognized as a time-consuming and monotonous task not everyone excels at, requiring a certain level of search skills. According to Google, the tool can reduce hours of research on almost any subject to just a few minutes.

Currently, Deep Research is available to users on the web version of Google Workspace, with mobile support expected to follow soon. It currently supports English only, but Google plans to expand to other languages shortly.

Additionally, Google has announced several new experimental models within Gemini Advanced, including 2.0 Flash Thinking Experimental and 2.0 Pro Experimental, which currently rank second and third globally in the chatbot arena's LLM leaderboard.

The 2.0 Flash Thinking Experimental model represents Google's most advanced "reasoning model" to date. Unlike traditional generative AI models, it breaks down prompts into a series of steps to enhance its reasoning capabilities and provide better responses. Uniquely, it outlines its entire thought process, allowing users to trace its reasoning and understand the assumptions it makes.