Google's AI offensive: new tools, new possibilities
With Gemini 2.0, Notebook LM, Vision AI, and other tools, Google is launching a new generation of AI tools. These tools promise greater efficiency, creativity, and personalization for developers and users alike.
At the latest "KI-ipfeli" media breakfast in Zurich, Google presented various innovations in the field of AI. Evelyn Leu, Lead Education Manager at Google Switzerland, hosted the event and emphasized: "Don't be afraid of it! Just try them out. These tools are often free and help you to make faster progress and find better solutions." The event showed that the race for supremacy in the field of AI has gained new momentum.
Gemini 2.0: The "versatile sparring partner"
The heart of the new Google AI platform is "Gemini 2.0". Leu explains: "I think Gemini is very good because it can support you in your work. You no longer have to start with a blank sheet of paper, you have a sparring partner to support you." The "Gems" function, which provides various AI personas, is particularly outstanding. Users can choose between a creative partner, career advisor or coding assistant, allowing them to personalize their work.
In addition to the standard version of Gemini 2.0, Google offers specialized versions:
- Gemini 2.0 Flash: This model is ideal for high-frequency and scalable tasks. It can process large amounts of data quickly and has a context window of 1 million tokens.
- Gemini 2.0 Flash Thinking Experimental: This version combines the speed of Flash with advanced capabilities to analyze more complex problems. It is particularly suitable for creative applications and strategic thinking.
- Gemini 2.0 Pro: The most powerful model for coding and complex queries. With a context window of 2 million tokens, it enables comprehensive analyses and can integrate external tools such as Google Search or code interpreters.
- Gemini 2.0 Flash-Lite: This version aims to post with high cost efficiency and improves performance compared to older models such as 1.5 Flash. It is suitable for price-conscious users.
Gemini 2.0 adapts to your needs, learns continuously and helps you discover new subject areas. Whether creative projects or complex analyses - Gemini is a flexible and efficient companion.
Notebook LM: Linking and deepening
Notebook LM was developed to combine and analyze information from different sources. Evelyn Leu emphasizes: "You can upload files, websites or videos, extract information and even create FAQs or summaries."
The strength of Notebook LM lies in its versatility. Users can upload materials from various sources such as PDFs, Google Docs, websites and even YouTube videos. The tool analyzes this data, creates links between them and prepares them in user-friendly formats such as timelines, learning aids or FAQs. The function to extract important information and create summaries that link directly to the original sources is particularly practical.
Notebook LM offers enormous advantages for journalists, researchers and creative projects. It not only enables the comparison and linking of content, but also personal customization. Users can ask specific questions and receive answers that are tailored to their individual needs. The variety of display options - from classic texts to audio summaries - makes Notebook LM an indispensable tool for anyone working with large amounts of information.
Document AI: structured data at the touch of a button
With Document AI companies can automate tedious document processing tasks. Evelyn Leu describes: "Document AI takes a lot of work off your hands. For example, you can take a photo of a document and upload it, and the tool automatically recognizes the relevant fields." Whether expense reports or bank documents - Document AI saves time and turns unstructured data into structured information. Click here for the demo version of Document AI.
Vision AI: recognizing and analysing patterns
With Vision AI Google relies on the automatic analysis of images. The areas of application range from document processing to tumor research. Evelyn Leu describes: "Vision AI analyses images, recognizes patterns or checks whether content is admissible." The tool makes it easier for companies and institutions to evaluate large visual data sets. From image recognition to medical applications - Vision AI saves time and delivers precise results. Click here for the demo version of Vision AI.
Imagen 3: Creative image generation
Imagen 3, Google's latest text-to-image model, takes image generation to a new level. It creates visually appealing content - from photorealistic landscapes to artistic styles. The improved color balance and level of detail make Image 3 a versatile tool for creatives and marketers.
Prompt better: Tips for effective requests
Evelyn Leu emphasized the importance of well-formulated prompts: "Think before you prompt. It helps to be specific and give clear guidelines." Here are her tips and here exclusively at m&k the link to the Gemini-Prompting-Guide.
- Use natural language.
- Be precise and give context.
- Iterate your prompts to achieve better results.
- Use the correct output formats.
- Personalize prompts with your own documents.
Energy efficiency: Think before you prompt
A crucial point when using AI is energy consumption. Evelyn Leu advises: "Think before you prompt. AI agents require more energy than traditional search queries." Nevertheless, Google is working on more efficient models to reduce energy consumption. However, if you want to perform simple tasks, you should continue to use classic search queries.
Conclusion: opportunities and challenges
Google entered the generative AI race relatively late. With tools such as Gemini 2.0, Notebook LM, Vision AI, Document AI and Imagen 3, the company now wants to catch up. Access to the free web in particular makes these tools an important companion for creatives, developers and companies. Evelyn Leu encourages: "Don't be afraid of it! Just try them out. These tools are often free and help you make progress faster and find better solutions."
