Google just introduced its super smart AI model, Gemini and it comes in different versions: Ultra, Pro, and Nano. Each version does specific things like Ultra handles complex tasks, Pro is for a bunch of different stuff, and Nano is for things on your device.
"Seeing some qs on what Gemini *is* (beyond the zodiac :). Best way to understand Gemini’s underlying amazing capabilities is to see them in action, take a look", stated Sundar Pichai, CEO of Alphabet, while sharing a video on X (Formerly Twitter)
So, Google's chatbot Bard is using Gemini Pro to do some fancy reasoning and planning. And guess what? It's not just for the AI world; it's in the new Pixel 8 Pro too. They're using it for features like Summarise in the Recorder app and Smart Reply in Gboard.
What's cool about Gemini is that it can understand all sorts of things—text, code, audio, images, videos. They say it's "multimodal," which basically means it's a jack of all trades.
According to the company, Gemini is like a superhero in the AI world, rocking tasks from understanding pictures and audio to doing complex math.
“With a score of 90 per cent, Gemini Ultra is the first model to outperform human experts on MMLU (massive multitask language understanding), which uses a combination of 57 subjects such as math, physics, history, law, medicine and ethics for testing both world knowledge and problem-solving abilities,” said Google.
Furthermore, it can actually write code too. It understands languages like Python, Java, C++, and Go. Imagine having an AI buddy that helps you with coding—that's Gemini.
Google trained Gemini using their special machines called Tensor Processing Units. They even launched a new one called Cloud TPU v5p, the big gun for training super-advanced AI.
The plan is to bring Gemini into more Google stuff, like Search, Ads, Chrome, and some AI thing called Duet. So, it's not just an AI thing; it's becoming a Google thing everywhere. It's like Google's stepping up its tech game big time!