Everything You Should Know About Google's Latest Released AI Model - Gemini Ultra
Google has recently unveiled its much-anticipated AI model, Gemini, which it claims to be more potent than OpenAI's GPT-4 large language model (LLM). According to Google, Gemini can compete at an "expert level" with human intelligence in an intelligence test.
During Google's I/O conference in May, CEO Sundar Pichai introduced Gemini, which will be available in three different versions - Nano, Pro, and Ultra. The Ultra variant is claimed to be the most proficient and can potentially outperform GPT-4.
What Is Gemini And Who Made It?
Google's latest artificial intelligence model Gemini is a robust and advanced system that can comprehend more than just text. It has the ability to recognize images, videos, and audio as well. Described as a multifaceted model, Gemini can perform intricate tasks in areas like mathematics and physics. Additionally, it can also create superior quality code in several programming languages.
Gemini can currently be accessed via Google Bard and Google Pixel 8, and its integration into other Google services will take place gradually over time.
Gemini is the result of large-scale collaborative efforts by teams across Google, including our colleagues at Google Research," according to Dennis Hassabis, CEO and co-founder of Google DeepMind. "It was built from the ground up to be multimodal, which means it can generalize and seamlessly understand, operate across, and combine different types of information including text, code, audio, image, and video."
Google and its parent company, $Alphabet-A(GOOGL.US)$, collaborated to build Gemini, which is the most sophisticated AI model released by the company thus far. The development of Gemini also saw significant contributions from Google DeepMind.
Gemini's Performance Compared to OpenAI's GPT-4
Google announced that its Gemini Ultra model has outperformed human experts in the massive multitask language understanding (MMLU) test, scoring 90% compared to the expert level's 89.8%. However, Google's rival OpenAI's GPT-4 scored only 86.4% on the same test. The mid-range Gemini Pro model surpassed only GPT-3.5, indicating a close competition between the two AI models.
Gemini is one of the most advanced models to date, with its true capabilities set to be determined by the release of the Ultra variant. What makes Gemini stand out from other AI chatbot models, including GPT-4, is its innate multimodal characteristic, unlike others that require plugins or integrations for such functionality.
From December 13, developers and enterprise customers can access Gemini Pro through the Gemini API in Google's AI Studio and Google Cloud Vertex AI. Android developers will have early preview access to Gemini Nano via AICore.
What Does It Mean for Google's Products?
• Bard to Utilize Gemini for Advanced Reasoning and Planning
Google also has announced a significant upgrade to Bard, integrating Gemini Pro to enhance the AI's capabilities.
With Gemini, Bard is getting its biggest upgrade yet," said Bard vice president Sissie Hsiao.
Gemini ramps up the quality of Bard's performance, whether in writing poetry or computer code to shopping queries or research projects, according to Hsiao.
Google in September integrated Gmail, YouTube and other tools into its Bard chatbot as tech giants seek to persuade users that generative AI is useful and not dangerous or just a fad. Those capabilities closely match offerings from Microsoft that infuse its Office 365 apps with AI powers, though those come at an extra cost to customers and are not available through the chatbot on its search engine Bing.
• Google Tests Gemini for Search Generative Experience (SGE)
We're already starting to experiment with Gemini in Search, where it's making our Search Generative Experience (SGE) faster for users, with a 40% reduction in latency in English in the U.S., alongside improvements in quality."
• Google Pixel 8 Pro: First Smartphone with Gemini Nano's Built-In AI
Google's latest update introduces Gemini Nano, an advanced AI model, now integrated into the Pixel 8 Pro smartphone.
Google's new update for Pixel 8 Pro includes advanced AI features that leverage Google Tensor G3 technology and Gemini Nano, making it the first phone designed for AI. The 'Summarize in Recorder' and 'Smart Reply in Gboard' features enhance user privacy without requiring a network connection. The update also improves photography and video with enhanced video stabilization, Night Sight video, and Photo Unblur for better pet images.
The Launch Has Had a Mixed Reception
The big deal is that it appears to be the first model to beat GPT-4," wrote Ethan Mollick, a professor at the University of Pennsylvania's Wharton School, on X. "The fascinating thing is that it does it by just a tiny bit."
Meanwhile, Jesse Dodge, a research scientist at the Allen Institute for AI, wrote that despite Google's claims that its training data for Gemini was key to its performance, the company provided "almost no information about how it was made, how it was filtered, or its contents."
Yesterday, $IBM Corp(IBM.US)$ and $Meta Platforms(META.US)$ announced the establishment of the AI Alliance, an open innovation coalition comprising over 50 organizations in the AI industry. Notably, early AI pioneers such as OpenAI, Google, and Amazon were absent from the alliance, which could suggest a strategy to limit their impact on the industry.
Though Google claims Gemini Ultra has the potential to surpass GPT-4, it's still not ready for mainstream use. Therefore, any definitive conclusions must be postponed until further development and testing.
Source: Google DeepMind, ZDNET, Futurism, Investopedia, Business Insider, France24, Search Engine Journal
Disclaimer: Moomoo Technologies Inc. is providing this content for information and educational use only. Read more