New Artificial Intelligence Model Google Gemma

Google continues its progress in the field of artificial intelligence unabated. Google Gemma, the next-gen artificial intelligence model, was officially announced.

Google Gemma is a focused solution for artificial intelligence developers and researchers, developed by Google DeepMind employees and other internal teams. Gemma, which stands out with its open and lightweight structure, attracts attention by bringing more than one model together. Currently, the "Gemma 2B" and "Gemma 7B" versions are available.

Google Gemma is designed to be able to work locally on laptops and desktop computers.  It is also compatible with IoT, mobile and cloud platforms. Gemma can be utilized for various processes, such as ChatGPT, and has a wide range of utilization areas.  Google plans to offer more Gemma models in the coming period and encourages Gemma for companies and independent developers to develop artificial intelligence-powered applications. Previously, Google had announced a product called Goose. Goose serves as an LLM, a major language model that speeds up the coding process, and increases the productivity of company employees and uses Gemini technology on its basis.

Goose, which was first revealed by Business Insider, was trained using Google's internal data and does not pose a data security risk as an integrated solution.  Many companies do not prefer their employees to use systems such as ChatGPT because they are concerned about the reliability of systems trained with user input.  Google explains that Goose can "answer questions about Google-specific technologies, generate code using internal technology know-how, and support capabilities such as natural language-based code editing." This attracts attention as part of a rapidly advancing trend in the field of artificial intelligence. The recently announced next-gen artificial intelligence model Gemini 1.5 is the next version of the rapidly developing Gemini series and is currently only available for testing by developers and corporate users.

Google plans to launch its new systems for everyone in the future period and promises to offer a much more comprehensive productive artificial intelligence experience with the 1.5 version prepared.  This version ameliorates the user experience by significantly increasing the response speed of reply. Google is also preparing version 1.5 Pro, which is said to be on the same level as Gemini Ultra. Gemini 1.5 has a large context window and is able to process larger queries and access more information at the same time. This considerably increases the capacity of the system. For comparison, it is stated that 128 thousand tokens can be used for OpenAI's GPT-4 model, 32 thousand for Gemini Pro, and a full 1 million "token" value for the Gemini 1.5 version.