Top-Tier Model Equipped 'Advanced' Paid Subscription Also Available
Google's generative artificial intelligence (AI) chatbot Bard is being renamed to 'Gemini.' Additionally, the top-tier model of Gemini, called 'Ultra,' which will compete with ChatGPT's 'GPT-4,' is being launched.
On the 8th (local time), Google announced plans to expand the Gemini ecosystem through its company blog. Gemini, unveiled last December, is a multimodal AI model that generates not only text but also images, voice, and video, and is categorized into Ultra, Pro, and Nano based on the scale of training.
On this day, Google announced that with the release of Gemini's top-tier model Ultra, it will unify the names of all its AI products under 'Gemini.' Accordingly, Google's AI chatbot, previously known as 'Bard,' will be renamed 'Gemini.' Furthermore, the AI features embedded in Google's collaboration software suite Workspace?including Google Docs, Gmail, Sheets, and Slides?known as 'Duet AI,' will also be renamed to Gemini.
Google explained, "(The name unification) expands the term to refer to the ecosystem encompassing AI-related products and services," adding, "It is intended to highlight Gemini's core multimodal capabilities beyond a simple model name."
On this day, Google launched the AI chatbot 'Gemini Advanced' equipped with the first version, Ultra 1.0. Ultra is a model that, when revealed last December, achieved a 90% accuracy rate on the Massive Multitask Language Understanding (MMLU) benchmark, which evaluates knowledge and problem-solving skills across 57 subjects including mathematics, physics, history, law, and medicine, surpassing human experts. Google stated that Ultra is the first model to exceed not only OpenAI's latest model GPT-4 (86.4%) but also human expert scores (89.8%). This multimodal chatbot also possesses coding, advanced reasoning, comprehension, and creative capabilities.
This chatbot is available in English starting today in over 150 countries including the United States and South Korea, with plans to expand the supported languages soon. Additionally, Google has released a dedicated app that allows easy access to Gemini on smartphones such as Android phones and iPhones. Currently, the app is available in English in the U.S., with expansions to Korean, Japanese, and other languages starting next week. However, the release date in Europe remains undecided due to regulatory issues.
'Gemini Advanced' can be accessed through the subscription service 'Google One AI Premium Plan,' which includes additional features such as Google Photos, Meet (video conferencing), and Drive (cloud services), at a monthly fee of $19.99 (approximately 26,000 KRW). Competitor OpenAI offers its paid subscription service ChatGPT Plus, based on GPT-4, at a nearly identical price of $20 per month.
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
