Google has announced the integration of its Bard chatbot with the cutting-edge artificial intelligence (AI) model known as Gemini. This new-generation AI is hailed by Google as a superior reasoning system compared to ChatGPT and other competitors. The unveiling of Gemini comes as Google aims to secure the lead in generative AI, challenging OpenAI’s ChatGPT.
Google has for years discreetly developed AI powers but was caught off guard when OpenAI released ChatGPT in collaboration with Microsoft late last year, making its advanced capabilities accessible globally, according to Agence France-Presse.
Google and Alphabet CEO Sundar Pichai in a release expressed excitement about the new era of models, stating, “This is incredible momentum, and yet, we’re only beginning to scratch the surface of what’s possible. This new era of models represents one of the biggest science and engineering efforts we’ve undertaken as a company.”
Pichai took to X (formerly known as Twitter) to announce Gemini 1.0, describing it as Google’s “most capable and general AI model yet.” Gemini is optimized in three sizes – Ultra, Pro, and Nano – catering to diverse needs and applications. Gemini Ultra, in particular, has achieved groundbreaking results, surpassing human experts on various academic benchmarks.
Demis Hassabis, CEO and Co-Founder of Google DeepMind in a statement highlighted the flexibility of Gemini, which efficiently runs on data centers and mobile devices alike. Gemini 1.0’s multimodal reasoning capabilities enable it to comprehend complex written and visual information simultaneously, making it adept at extracting insights from hundreds of thousands of documents through reading, filtering and understanding information.
The AI model’s proficiency extends to understanding text, images, audio, and more, enabling it to answer questions on intricate subjects. Gemini 1.0’s advanced coding abilities set it apart, as it can understand, explain, and generate high-quality code in popular programming languages such as Python, Java, C++, and Go. Gemini Ultra’s excellence in coding benchmarks, including HumanEval, positions it as a leading foundation model for coding globally among many others.
During a briefing, Eli Collins, the Vice President of Product at Google DeepMind, announced that Gemini is the first AI model to surpass human experts in specific benchmarks related to problem-solving, math, physics, history, law, medicine, and ethics.
A demonstration showcased Gemini’s ability to recognize various inputs, from a person reenacting a scene from the “Matrix” movie to someone drawing a duck and holding up a rubber duck. Gemini not only identified these elements but also provided comments, comparisons, drew conclusions, and offered suggestions, Agence France-Presse reported.
According to Collins, the performance of the “Ultra” version of Gemini significantly surpasses that of other state-of-the-art models in 30 benchmark tests, measuring capabilities such as image understanding and mathematical reasoning.
Sissie Hsiao, the Vice President of Bard, stated, “With Gemini, Bard is getting its biggest upgrade yet.” Bard will leverage Gemini for more advanced reasoning, planning, and understanding capabilities, as demonstrated during the briefing.
Gemini is set to be available in English in over 170 countries and territories, with additional language support to be added soon, according to Hsiao. She added, “With Gemini, we are one step closer to our vision of bringing you the best AI collaborator in the world.” Hsiao emphasized that Gemini elevates Bard’s performance across various tasks, including writing poetry or code, answering shopping queries, and conducting research projects.
