Google has confirmed the launch of its latest, most advanced artificial intelligence (AI) model, Gemini, which will be available in three different formats.
From data presented by the tech giant, Gemini appears to be the first model to be able to outperform GPT-4 in many areas.
The official release issued today by Google stated, “Gemini is also our most flexible model yet — able to efficiently run on everything from data centers to mobile devices.”
“Its state-of-the-art capabilities will significantly enhance the way developers and enterprise customers build and scale with AI.”
As stated, it comes in three different sizes, depending on the type of task required. Gemini Ultra is the largest offering and also the most capable model for highly complex needs. Gemini Pro is the median offering with a wide scope and Gemini Nano is the compact version, best suited for on-device tasks.
Crucially, the power of Gemini 1.0 is reflected in its ability to run on everything, from extensive data centers to the mobile device that is in your pocket.
Another serious asset for this Google launch is that Gemini is the first model to outperform human experts on MMLU (Massive Multitask Language Understanding). It is one of the most popular methods to test the knowledge and problem-solving abilities of AI models, using a combination of 57 subjects such as math, physics, history, law, medicine, and ethics for testing both world knowledge and problem-solving abilities.
Google’s Gemini is ‘next step on our journey’
Sundar Pichai, Google CEO said of the landmark launch, “I believe the transition we are seeing right now with AI will be the most profound in our lifetimes, far bigger than the shift to mobile or the web before it.”
“AI has the potential to create opportunities, from the every day to the extraordinary, for people everywhere.”
“Now, we’re taking the next step on our journey with Gemini, our most capable and general model yet, with state-of-the-art performance across many leading benchmarks.”
Google plans to license Gemini to users through Google Cloud to use in their applications, whilst it will power consumer-facing Google AI apps like the Bard chatbot and Search Generative Experience.
This is a bold, headline move from the tech giant as it aims to go head-to-head with OpenAI, the makers of ChatGPT in what is a rapidly fluid and competitive playing field.
However, it will not yet be available in Europe or the UK, which Google suggested was down to regulatory hurdles, with more information to follow.
This all comes the day after Microsoft announced it was to supercharge its own Copilot with OpenAI upgrades.
Image credit: blog.google
