Google Unveils Its Latest and Greatest Gemini AI, Beating ChatGPT

Google has just introduced Gemini, a groundbreaking AI model described as their most advanced and versatile yet. It is primarily meant to upgrade the AI chatbot Google Bard.

Ideal for Programming and Education

Designed from the ground up to be multimodal, Gemini excels in processing and interpreting a variety of information forms simultaneously, including text, images, audio, video, and code. This comprehensive approach enhances its ability to grasp subtle details, significantly improving its capability to tackle complex inquiries. For that reason, Gemini particularly shines in providing aid in more complex fields such as mathematics and physics, offering clear explanations and reasoning.

Gemini comes in three distinct versions: Nano, Pro, and Ultra each tailored to work with different levels of complex tasks. Nano, the lowest tier is described as the “most efficient model for on-device tasks”, probably suggesting that it is meant to respond quickly to simpler queries.

In contrast, the Pro version is designed as “the “best model for scaling across a wide range of tasks”, and finally, the highest tier Ultra is “the largest and most capable model for highly complex tasks.”

The Ultra variant of Gemini particularly stands out in its ability to comprehend, articulate, and produce high-quality code in multiple programming languages such as Python, Java, C++, and Go. Its proficiency extends across different languages, allowing it to navigate and rationalize complex information seamlessly.

This model demonstrates exceptional performance in various coding benchmarks, including HumanEval and Natural2Code. Notably, Natural2Code is a proprietary Google dataset that prioritizes original, author-generated content over web-based sources.

Gemini Pro is Available in Bard

Google’s Bard AI chatbot has been enhanced with a “fine-tuned version” of Gemini Pro. This upgrade is expected to significantly improve its capabilities in advanced reasoning, planning, and understanding, according to Google.

Remarkably, Gemini Pro has reportedly surpassed GPT-3.5 in six of the eight benchmarks conducted by Google. Additionally, in blind evaluations conducted by independent third-party raters, Bard, powered by Gemini Pro, has been recognized as the most preferred free chatbot when compared to other leading options, although Google hasn’t specified which alternatives were included in this comparison.

Currently, Bard with Gemini Pro is available for text-based prompts in English across 170 territories, excluding Europe, where it is expected to be launched soon. Furthermore, Google plans to introduce Gemini Ultra to the Bard Advanced experience early next year.

Google Pixel 8 Pro has Gemini Nano

In a related development, the Pixel 8 Pro has been announced as the first smartphone specifically designed to incorporate Gemini Nano. This integration is not extended to the standard Pixel 8. Notably, features such as Summarize in Recorder and Smart Reply in Gboard on the Pixel 8 Pro are already utilizing Gemini Nano.

Google is also beginning to integrate Gemini into its Search service to enhance speed and efficiency. In the upcoming months, Gemini is set to power various features in more Google products and services, including Ads, Chrome, and Duet AI.