What's with all this buzz around Google Gemini?
Gemini is the latest Large Language Model (LLM) from Google that will power the AI capabilities of several Google products and is already replacing PaLM2 on Bard.
On December 6, Alphabet released the first phase of its next-generation AI model, Gemini.
Google Gemini represents a step forward for Google’s AI efforts, as it is a brand-new large language model (LLM) made by Alphabet’s Brain Team and DeepMind.
Gemini can generate and process text, images, and other kinds of data like graphs and maps. That’s right — the future of AI isn’t just chatbots or image generators.
How Gemini compares with GPT 4 from OpenAI?
The image added, represents the clear winner when it comes about gemini.
Gemini is also the first model to outperform human experts on MMLU (Massive Multitask Language Understanding), one of the most popular methods to test the performance of language models.
There are 3 versions of Gemini
1. Gemini Nano: doesn't have a direct comparison with GPT4 and is the most lightweight and efficient of the models, designed to run directly on mobile devices. Google's upcoming Gemini Nano is set to debut as a preview feature within the new AI Core application, initially rolling out to Pixel 8 Pro users beginning December 6. While it's anticipated to expand to additional Android 14 devices in the future, currently, this AI technology is exclusive to Pixel users.
2. Gemini Pro: this is the engine that runs the upgraded Bard experience. Google claims that Gemini Pro is more capable than GPT-3.5 in six different benchmarks and is specially optimized for tasks like brainstorming, summarizing content, and writing. It was launched on December 6th as an under-the-hood upgrade to Google Bard. It is also expected to roll out to enterprise customers using Vertex AI on December 13.
3. Gemini Ultra: The Ultra exceeds 30 of 32 academic benchmarks for current state-of-the-art results used for LLMs and beats GPT-4 in every category outside of commonsense reasoning for everyday tasks.
Unlike GPT-4, which processes only words and images, Gemini advances further by comprehending a broader spectrum, including words, images, audio, coding, as well as complex subjects in mathematics and physics. It boasts the ability to respond to queries almost instantaneously, offering real-time interaction. However, Gemini is currently not accessible in any form, rendering it more of a prospective offering than a present reality.