Gemini Advanced: Multimodal AI chatbot
The fourth part of our webinar series focused on Gemini Advanced. We demonstrated how the multimodal AI chatbot confidently masters challenging tasks while seamlessly integrating into Google Workspace.
We summarize the most important findings here:
1. What makes Gemini Advanced special?
Gemini Advanced is much more than a simple chatbot. Powered by Google's most powerful models, it offers a wide range of benefits:
- Multimodality & reasoning: AI can not only read texts, documents, and data, but also analyze them, understand complex relationships (reasoning), and draw logical conclusions.
- Performance: Thanks to continuous model advancements, AI delivers faster responses, processes larger language inputs, and significantly reduces so-called "hallucinations" (misinformation).
- Workspace integration: Gemini operates directly in your familiar environment. It can launch specific prompts on Google Workspace instances and evaluate documents without having to switch applications.
Safety comes first
Data protection is a crucial issue for companies. Since Gemini Advanced is an extension of Google Workspace, the same enterprise security settings apply as in the rest of Workspace.
Important: Google does not use your interactions and company data to train the models.
2. Gems: Personal AI assistants
A highlight of the webinar was the presentation of the "Gems, " which can be trained for specific topics or recurring tasks.
- Customized workflows: You can create your own gems that are tailored to the communication style of a department or to defined processes.
- Predefined or custom: You can use predefined gems from Google (such as the "Brainstormer" or experimental tools such as the "Storybook" creator) or develop your own, specifically customized variants.
- Teamwork: Created Gems can be shared, edited, or copied within the organization, which facilitates knowledge management within the team.
3. Use cases: Gemini in practice
Examples of how different departments can use Gemini:
- Marketing: Target group analysis, market research, and creation of campaign drafts.
- Sales: Preparation of client meetings (client discovery) and strategy development.
- HR: Support with recruiting and creating onboarding materials.
- Finance: Analysis of complex data sets and creation of financial models.
- Software development: code generation, quality assurance (QA), and testing.
4. Deep Dive: Canvas: The new way of editing text
The Canvas feature is a special workspace within the Gemini application that feels like a Google Doc but is AI-powered.
The top features of Canvas:
- Direct editing: Unlike in traditional chat, texts can be edited directly here. If a change is desired, Gemini does not regenerate the entire text, but only adjusts the desired sections.
- Slider tools: Texts can be shortened or lengthened using intuitive sliders.
- Seamless export: Content (including headings and formatting) can be exported directly to Google Docs or Google Slides.
- Code integration: Canvas is suitable for creating scripts and code. Questions about individual lines of code can be asked directly, and the code can also be optimized in the window.
Example from the live demo: Complex presentations can be created based on uploaded sales figures (Google Sheets). These can be exported directly to Google Slides with image source references and finalized there.
Conclusion: The proof of the pudding is in the eating.
The webinar made it clear that the tools are there to massively increase productivity and creativity. It is worth taking the time to test features such as Gems and Canvas. The output often improves through repetition and trying out different models.
