Google is making updates to its AI-powered chatbot, Gemini, in order to keep up with its competitors in the generative AI space. The updates are aimed at improving the performance of the platform and making it more accessible to users. One of the major updates is the release of Gemini 1.5 Flash, a lightweight multimodal model that will be available on the web and mobile in 40 languages and around 230 countries.
Google claims that Gemini 1.5 Flash delivers upgrades in quality and latency, with significant improvements in reasoning and image understanding. The model is also more cost-effective to run on the backend, which is a major advantage for Google. By reducing costs, Google can provide a better user experience while also saving money.
In addition to the new base model, Google is expanding Gemini’s context window to 32,000 tokens. This allows the model to consider a larger amount of input data before generating output. Models with larger contexts have several advantages, such as the ability to summarize and reason over longer text snippets and files. In a chatbot context, larger contexts also reduce the likelihood of the model forgetting topics that were recently discussed.
Another update to Gemini is the ability to upload files for analysis. Previously, this feature was only available in the paid edition of Gemini, but Google is now enabling file uploads from Google Drive and local devices for all users. This means that users can upload files and ask Gemini to create practice questions or analyze data files, allowing them to uncover insights and visualize them through charts and graphics.
To address the issue of hallucinations, where the model generates false information, Google is previewing a feature that displays links to related web content beneath certain Gemini-generated answers. This provides users with additional sources of information and helps them verify the accuracy of the model’s responses. This feature is part of Google’s efforts to make the AI more transparent and reliable.
Google is expanding the availability of Gemini in various ways. It is rolling out the feature in Messages for select devices in the European Economic Area, U.K., and Switzerland, with support for additional languages. Google is also launching the Gemini mobile app in more countries and expanding access to teenagers globally. The company has introduced a teen-focused Gemini experience, allowing students to sign up using their school accounts. To protect teens and ensure responsible use of AI, Google is implementing additional policies and safeguards, as well as providing an AI literacy guide.
Overall, these updates to Gemini demonstrate Google’s commitment to improving its AI chatbot and making it more accessible to users. By enhancing performance, expanding language support, and introducing new features, Google aims to provide a more reliable and user-friendly experience while maintaining its competitiveness in the generative AI space.
