Google began the phased rollout of its latest artificial intelligence software, Genesis, on Wednesday, in a highly-anticipated release as the AI race heats up.
Genesis 1.0, the company’s first version, has been optimized into three different sizes: Gemini Ultra, the largest and most capable of highly complex tasks; Gemini Pro, best for scaling across a wide range of tasks; and Gemini Nano, the most efficient for on-device tasks.
Google’s chatbot Bard is now using Gemini Pro for advanced reasoning and understanding, which the company says is Bard’s greatest upgrade since its launch. The souped-up version of Bard is currently available in English in more than 170 countries, and the company plans to expand to other languages and locations soon.
META, IBM LAUNCH INTERNATIONAL ‘AI ALLIANCE’ WITH 50 ORGANIZATIONS TO ADVANCE OPEN-SOURCE AI
Sissie Hsiao, Google’s vice president and general manager of Bard and Google Assistant, wrote in a blog post that in testing prior to its public debut, Gemini Pro outperformed ChatGPT developer OpenAI’s GPT-3.5 in six out of eight benchmarks used for measuring large AI models, including Massive Multitask Language Understanding (MMLU).
Also on Wednesday, Gemini Nano is powering new features on Google’s Pixel 8 Pro smartphone, including a “Summarize” feature in the Recorder app, and “Smart Reply” in Gboard, starting with WhatsApp.
STUDY FINDS CHATGPT PROVIDED INACCURATE ANSWERS TO MEDICATION QUESTIONS
Beginning on Dec. 13, developers and enterprise customers can access Gemini Pro through the Gemini API in Google AI Studio or Google Cloud Vertex AI, the tech giant said.
Google did not give a date for when Gemini Ultra – its answer to Microsoft-backed OpenAI’s GPT-4 – will become widely available, but said it will be rolled out to developers and enterprise customers early next year.
GET FOX BUSINESS ON THE GO BY CLICKING HERE
According to Google, Gemini Ultra is the first large language model to outperform human experts on MMLU, with a score of 90%. The company said OpenAI’s GPT-4 reportedly scored 86.4% on the measure.