Unveiling Our Ultimate Achievement: The Most Capable Model in Our History

OpenAI’s leading position in the generative AI space may be in jeopardy as Google unveiled its most powerful large language model yet, named Gemini, on Wednesday. This marks the beginning of “a new generation of AI models, inspired by the way people understand and interact with the world,” according to CEO Sundar Pichai.

Pichai explained, “Ever since programming AI for computer games as a teenager, and throughout my years as a neuroscience researcher trying to understand the workings of the brain, I’ve always believed that if we could build smarter machines, we could harness them to benefit humanity in incredible ways.”

Developed by Google’s DeepMind and Research divisions, Gemini boasts cutting-edge genAI capabilities. The system has been designed as an integrated multimodal AI, enabling it to understand and reason about various inputs more effectively than existing multimodal models. This allows Gemini to provide better responses on challenging subjects, such as physics.

Gemini is also proficient in popular programming languages and has been used to create AlphaCode 2, which outperformed its predecessor by solving twice as many challenge questions. Additionally, the model’s operational flexibility and ability to work in different form factors make it suitable for use in large data centers to local mobile devices.

Google has announced that Gemini will be available in three sizes: Nano, Pro and Ultra. Nano is designed for on-device tasks, while Pro is more versatile and will soon be integrated into Google’s existing products, including Bard. A specially-tuned version of Pro will be used to enhance the Bard chatbot, with plans to expand its availability in the coming years. Gemini Ultra, the most powerful version, is slated for release in 2024.

Furthermore, Pro’s capabilities will be accessible via API calls through Google AI Studio or Google Cloud Vertex AI, and Gemini functionality will be integrated into various Google features in the coming months.

Gemini Ultra is expected to be released in 2024 and will undergo testing and feedback from select customers, developers, partners, and safety and responsibility experts. It promises to greatly advance AI development.