Alphabet, the parent company of Google, has unveiled its most advanced artificial intelligence model to date, named Gemini. This new AI model is capable of processing various forms of information, including video, audio, and text. Gemini represents a significant advancement over Google's previous AI technologies, offering more sophisticated reasoning and a greater degree of nuance in understanding information.
Alphabet CEO Sundar Pichai described Gemini as a major science and engineering effort undertaken by the company. Following the launch of OpenAI's ChatGPT, Google has been focused on developing AI software to rival Microsoft-backed AI technologies. A portion of the Gemini model technology has already been incorporated into Google's AI assistant Bard, with plans to release an advanced version through Bard in the near future.
Alphabet has developed three versions of Gemini, each designed for different levels of processing power. The most powerful version is intended for data centers, while the smallest version is optimized for mobile devices. The new model is not only more capable but also more efficient, requiring less cost to serve to users than previous models. Additionally, Alphabet announced a new generation of custom-built AI chips, the Cloud TPU v5p, designed to train large AI models more efficiently.
