Google has made a significant leap in the field of artificial intelligence with the launch of Gemini, its latest AI model. On December 6, Google CEO Sundar Pichai and Google DeepMind CEO and co-founder Demis Hassabis announced this development in a company blog post, positioning Gemini as a model that surpasses the capabilities of OpenAI’s GPT-4.
Gemini stands out for its optimization across different sizes and use cases, including Ultra, Pro, and Nano versions. This multimodal AI model has been designed to understand and integrate various types of information, setting a new standard in the industry.
Advancements and Applications of Gemini
One of the most notable advancements of Gemini is its proficiency in mathematics and specialized coding, areas where it reportedly outperforms GPT-4. Google claims that the Ultra version of Gemini achieves "state-of-the-art performance" in 30 out of 32 academic benchmarks commonly used in large language model development.
Furthermore, it impressively scores 90% on the massive multitask language understanding (MMLU) test, surpassing human expert performance. Jeff Dean, Google's chief scientist, highlighted that Gemini Ultra is the first model to achieve such a high score across 57 subjects on the MMLU.
The model's design allows it to seamlessly reason across different formats, including text, images, audio, and video. "We designed Gemini to be multimodal from the start," Dean said, emphasizing the integrated approach of the model compared to its predecessors.
In addition to its multimodal capabilities, Gemini boasts advanced programming skills. It can generate high-quality code using AlphaCode 2, an advanced code-generation system, and is capable of solving complex programming challenges and collaborating with developers.
Looking ahead, Google plans to expand Gemini's functionalities. "This is the biggest upgrade to Bard since it launched. It will be available in English in more than 170 countries and territories, and we plan to expand to different modalities and support new languages and locations in the near future," the company stated.
Moreover, Gemini is set to make its debut on the consumer market with its integration into Google's flagship phone, the Pixel 8 Pro. "Pixel 8 Pro is the first smartphone engineered to run Gemini Nano," the company announced, revealing that the model will power new features in the Recorder app and Smart Reply in Gboard, starting with WhatsApp.