Google is constantly adding new features to its Google Bard chatbot. However, it did not seriously compete with ChatGPT and other AI chatbots equipped with GPT-3 and GPT-4. Now, Google has announced that Bard is now powered by Gemini Pro, a new AI model from Google.
According to Google, Gemini is the most powerful artificial intelligence model the company has ever created. Designed to work in a variety of systems, from data centers to mobile devices, it can seamlessly process and integrate text, code, audio, images, and video, making it a universal AI model.
Google Gemini 1.0 can understand complex information and answer related questions with its multimodal capabilities, and it comes in three different sizes: Gemini Ultra for complex tasks, Gemini Pro for universal use, and Gemini Nano for simplified device operations.
Google Bard received the biggest update as it has now been upgraded to the Gemini Pro version, which means it can perform more advanced reasoning, planning, insights, and so on. It will be available in English in over 170 countries and territories. Pixel 8 Pro will soon receive an update to work with Gemini Nano, which will add features such as Summarize in the Recorder app, Smart Reply in Gboard and WhatsApp.
As for Gemini Ultra, it can perform a wide range of tasks, including natural understanding of images, audio and video, mathematical reasoning, and multimodal problem solving. According to Google, it outperformed human experts in the MMLU test, which assesses both knowledge and problem-solving abilities, and achieved an exceptional result in the MMMU test, a test of multimodal thinking. In addition, Ultra outperformed previous models in image recognition tests without the need for object symbol recognition, demonstrating its own multimodal capability and ability to perform complex reasoning.
Google plans to make Gemini Ultra available to a select group of users for early testing and feedback before rolling out to developers and enterprise customers in early 2024. In addition, in early 2024, Bard Advanced, a new artificial intelligence technology based on Gemini Ultra, will be launched, providing users with access to the most advanced capabilities.
In addition, Google introduced Cloud TPU v5p, the most powerful TPU system to date, which enables faster training of large-scale generative AI models and accelerates the development of new products and capabilities. To protect against harmful content, Gemini uses special security classifiers to detect, label, and filter harmful content such as violence and negative stereotypes. This multi-layered approach ensures that Gemini is safe and inclusive for all users.