Google, the world’s most popular search engine company, has just introduced its new AI system called Gemini.
This comes after a long period of waiting and testing. Google already offers the option to use Bard AI,
and now, with Gemini, OpenAI’s ChatGPT, a widely used generative AI, becomes a competitor. The company claims that Gemini surpasses humans in many aspects.
According to Google CEO Sundar Pichai, Gemini represents one of the most significant engineering and science-related products developed by Google in its history.
Pichai mentioned that Google has been working on AI for about 8 years, and Gemini has been meticulously prepared through extensive testing and improvements.
The notable feature of Gemini is its ability to function not only on high-end devices but also on smartphones.
Gemini AI operates in three modes: Ultra, Pro, and Nano. In Ultra mode, it utilizes the largest Large Language Model (LLM) for AI tasks.
Pro mode works with smaller LLMs, and Nano mode operates with the smallest LLMs, allowing Gemini to run locally on computers and phones.
This new Gemini AI puts Google in direct competition with other AI systems, particularly OpenAI’s ChatGPT.
Despite being an early investor in DeepMind and an initial player in the AI field, Google faced some lag in innovation compared to products like ChatGPT.
It remains to be seen if Gemini AI can help Google catch up with or surpass these newer AI tools.
According to Google DeepMind CEO and co-founder Demis Hassabis, Gemini has demonstrated superior performance compared to humans in various language tasks.
In a blog post, he mentioned, “Gemini Ultra achieves a score of 90.0 percent, making it the first model to surpass human experts in Massive Multitask Language Understanding (MMLU).”
MMLU is a comprehensive measure encompassing 57 subjects such as Mathematics, Physics, History, Law, Medicine, and Ethics, evaluating understanding of the world and problem-solving ability.
