Google Launches New AI, Gemini, and a More Powerful Bard Chatbot – The Messenger
Google on Wednesday debuted a new language language model, Gemini, that it hopes can compete with OpenAI’s industry-leading GPT-4.
Ahead, Gemini will power Google’s Bard chatbot, its rival to OpenAI’s ChatGPT. Google has made three versions of Gemini — Nano, Pro and Ultra — and while it has rolled out Nano and Pro across its services, it plans to debut Ultra, the most advanced version, later next year.
The Gemini launch comes amid a dramatic AI arms race in Silicon Valley, with Google and other competitors hoping to capitalize on the recent leadership woes at OpenAI.
For Gemini, Google has touted its ability to solve or check problems in complex fields like math and physics, generate advanced computer code and more far afield tasks, like the ability to recognize two different balls of yarn and then offer knitting advice.
Gemini’s basic versions work like ChatGPT, with a user interface built around search engine-type written prompts. But more advanced Gemini versions can interact with video, images and audio — making suggestions or tweaks without a text input. During a video demo, an artist drew instruments on a piece of paper, which Gemini recognized and used to create a song. As the artist sketched additional instruments, Gemini added them to the composition.
Gemini’s release comes a day after Microsoft announced improvements to its AI product, Copilot, and closely follows Anthropic’s release of a refined model of its chatbot, Claude, and OpenAI adding audio features to ChatGPT.
