Google Launches Gemini AI, Its Competitor To GPT-4
Google has introduced its latest AI model, Gemini, which is set to compete with OpenAI’s GPT-4. Unveiled during the I/O developer conference and launched in three different sizes, Gemini has outperformed GPT-4 in several benchmarks, according to Google. This new AI model is designed to eventually integrate into Google’s search engine, Chrome browser, and other platforms.
The first version of Gemini is called Gemini Nano, which is a lightweight model intended to run natively and offline on Android devices. Initially compatible with the Tensor 3 chip, Google aims to expand its compatibility to other smartphone models as Qualcomm, Samsung, and MediaTek develop chips capable of running Nano. The Pixel 8 Pro will be the first device to feature Gemini Nano, offering auto-summarization in the Recorder app and Smart Reply on Gboard.
Gemini Pro, a more powerful and energy-efficient model, will serve as the backbone for Google’s AI services and is already used by the Bard chatbot. Starting from December 13, developers and enterprise users will be able to access Gemini Pro through Google Cloud’s Generative AI Studio or Vertex AI.
The top-tier AI model, Gemini Ultra, is a multimodal model capable of working with video, images, and audio. Unlike the lower-end models, Gemini Ultra can process inputs and outputs beyond text. Google DeepMind CEO, Demis Hassabis, stated that Gemini Ultra will become even more versatile as it gains more sensory capabilities. Developers and enterprise clients can expect to have access to Gemini Ultra and Bard Advanced early next year.
In a comparison with GPT-4, Google claims that Gemini Ultra surpasses OpenAI’s model in 30 out of 32 academic benchmarks used in large language model research. The company also emphasizes that Gemini’s efficiency is due to its training on Google’s in-house Tensor Processing Units.
At the moment, Gemini only supports English. However, Sundar Pichai, CEO of Google, announced plans to expand Gemini’s language support to other languages soon. Additionally, Pichai revealed that Gemini will be integrated into Google’s search engine and Chrome browser, resulting in faster performance and improved user experience globally.
With Gemini’s entrance into the AI landscape, Google aims to establish itself as a prominent player, competing with the likes of OpenAI. The launch of Gemini, along with its future advancements and integrations, signifies Google’s commitment to pushing the boundaries of AI technology and enhancing various digital platforms and services.
