Google has unveiled its new AI model, Gemini. Designed to behave like humans, this model outperforms other models in tasks such as understanding, summarizing, reasoning, coding, and planning. Gemini comes in three versions: Pro, Ultra, and Nano. The Pro version is already available, with the Ultra version set to be released early next year.
Availability of Gemini:
Currently, Google has integrated the new Gemini Pro with its Chatbot boards, making it a direct competitor to ChatGPT. You can engage in text-based conversations with the Gemini-operated board, and Google promises support for other modes of interaction “soon.” While the new update is available in 170 countries and regions, its language support is limited to English.
Gemini and Google’s DeepMind Division:
Gemini, developed by Google’s DeepMind division, is a large language model (LLM) designed to compete with other AI systems like OpenAI’s ChatGPT, potentially surpassing them in performance.
Key Features of Google Gemini:
1. Multimodal Capabilities: Gemini is designed to integrate text, images, and other data types for more natural conversations.
2. Tools and API Usage: Gemini is part of Google’s new AI infrastructure, utilizing Pathways, indicating it might be the largest language model created so far.
3. Various Sizes and Capacities: Gemini comes in a “model scale” that offers different sizes and capacities, enhancing accuracy and reducing misinformation by learning from diverse sources.
Impact of Gemini:
The hope is that Gemini will have a significant impact on the AI industry. As Google’s most powerful AI model to date, it outperforms GPT-4 from OpenAI. It empowers applications and devices like board chatbots and Pixel 8 Pro. Google claims it is among the first models created as a multimodal LLM from the ground up, aiming to make conversations more natural and “human-like.”