GEMINI - SCI & TECH

News: Google Gemini - an AI model

 

What's in the news?

       Google has launched Project Gemini, its largest and most capable AI model.

 

About

       It is a Multimodal AI and a Large Language Model (LLM) developed by Google.

 

Features:

       Multimodal:

       It can understand different types of information including text, audio, images and video.

       Powerful:

       It surpasses most human experts in various tasks, potentially making it the most powerful AI model to date.

       Types:

       It will be available in three models:

1.      Ultra (for highly complex tasks)

2.      Pro (for scaling across a wide range of tasks)

3.      Nano (on-device tasks).

       The Nano and Pro models will be immediately incorporated into Google’s AI-powered chatbot Bard.

       The Ultra model will be launched next year.

 

Gemini Ultra:

       It is Google’s most powerful LLM ever and is aimed at enterprise applications that will run it for “highly complex tasks.

 

Gemini Pro:

       It is the most general-purposed of the three and has already been plugged into Bard for prompts that require advanced reasoning, planning and understanding.

 

Gemini Nano:

       It is described as the most efficient model for on-device tasks, has been baked into the Pixel 8 Pro to process tasks like information summarisation and Smart Reply.