As competition in the field of generative AI becomes increasingly fierce, Google has upgraded its AI chatbot Gemini to improve performance and expand its application scope. This update covers multiple aspects such as multi-language support, context window expansion, file upload functions, and transparency improvements, striving to maintain a leading position in the competition with giants such as Anthropic and OpenAI. The editor of Downcodes will give you an in-depth understanding of the latest upgrades of Gemini 1.5, as well as Google’s latest strategic layout in the field of AI.
Amid fierce competition with generative AI giants like Anthropic and OpenAI, Google has announced a comprehensive upgrade of its AI chatbot Gemini. This update focuses on improving the platform’s performance and expanding its usability, so let’s take a look at the important changes.
Multi-language support, global coverage
Starting this Thursday, Gemini 1.5 Flash (the lightweight multi-modal model released by Google in May) will support 40 languages on the web and mobile devices, covering approximately 230 countries and regions. Google said that Gemini 1.5 Flash has significantly improved quality and response speed, especially in reasoning and image understanding.
This upgrade not only improves user experience, but may also reduce Google's operating costs. Gemini1.5Flash is described as a streamlined and efficient version of Gemini1.5Pro, designed for narrowband, high-frequency generative AI workloads.
Source: google
Context window extension, file upload function
In addition to the new base model, Google has expanded Gemini's context window to 32,000 tokens, which equates to about 24,000 words (or 48 pages of text). This means Gemini can handle longer texts and provide a more coherent conversational experience.
What’s more noteworthy is that Google will soon open the ability to upload files from Google Drive and local devices to all Gemini users. Users can upload study materials, have Gemini create practice questions, and even analyze data files and generate visual charts.
Increase transparency and reduce illusions
To combat the possibility of AI models creating hallucinations (i.e. fabricating false information), Google is testing a new feature that displays links to related web content beneath certain answers generated by Gemini. This initiative aims to increase the reliability and transparency of information.
Wider application scenarios
Google is expanding Gemini capabilities into more areas:
1. Messages application integration: Gemini functionality will be launched in the European Economic Area (EEA), the United Kingdom and Switzerland, and support new languages such as French, Polish and Spanish.
2. Mobile application promotion: Gemini mobile application will be launched in more countries.
3. Open to teenagers: Gemini will be open to teenagers globally and will launch a dedicated onboarding process and AI literacy guide.
Google’s comprehensive upgrade of Gemini reflects its ambition and determination in the field of AI. By improving performance, expanding language support, adding new features and opening it up to a wider user base, Google is working to stay at the forefront of the highly competitive AI market.
All in all, Gemini’s upgrade demonstrates Google’s continued investment and technical strength in the field of AI. Its new features such as multi-language support, context expansion, and file upload will greatly enhance the user experience and expand its influence globally. Gemini's application scenarios will be further expanded in the future, which is worth looking forward to.