Google Unveils New AI Model, Gemini, to Compete with GPT Models
Google has announced the launch of its most ambitious artificial intelligence (AI) model, Gemini, in an effort to compete with the rapidly growing field of generative AI. The company stated that Gemini will be used to power everything from consumer apps to Android smartphones, marking a significant step forward in Google’s AI capabilities.
Unlike existing AI models that specialize in a single type of user request, Gemini is designed to be “multimodal,” accepting input that includes text, images, audio, video, and programming code. Google CEO Sundar Pichai described Gemini as a significant engineering effort for the company, emphasizing its potential to revolutionize AI capabilities in various contexts.
AI chatbot Bard has already been updated with a version of the Gemini model, and Google plans to integrate Gemini into widely used products such as the search engine and Chrome web browser. The announcement serves as an effort by Google to regain the lead in the AI race after being caught off guard by the success of OpenAI’s ChatGPT last year.
Gemini is available in three different sizes: Nano, Pro, and Ultra, each catered to different tasks and client needs. The launch also showcased Google’s advancements in cloud computing, with the company training Gemini using powerful cloud-based processors that can train large AI models three times faster than before.
In tests, Google’s Gemini model outperformed rival AI models in over two dozen benchmark tests, demonstrating its reading comprehension, math ability, and multi-step reasoning skills. However, Google has acknowledged the risks associated with AI models, including the potential for providing misleading results and hallucinating facts. The company has implemented additional techniques to improve the accuracy of the model’s responses.
Google plans to roll out Gemini Ultra, the most advanced version of the model, to select customers, developers, partners, and security experts for early experimentation and feedback. The model is undergoing third-party security assessments in alignment with Google’s commitment to the Biden Administration.
The launch of Gemini represents a significant step forward for Google in the AI sector, solidifying the company’s position as a key player in the industry. The company’s commitment to addressing potential risks associated with AI models demonstrates a proactive approach to ensuring the model’s accuracy and reliability.