Hajra Naz
According to The Information, Microsoft is developing a large-scale artificial intelligence (AI) language model to compete with Google and OpenAI.
According to a source claiming information from two Microsoft workers familiar with the project, Mustafa Suleyman, co-founder of Google DeepMind and previous CEO of AI firm Inflection, is supervising the model.
According to the sources, the tech giant may unveil the new model later this month at its Build developer conference. The model’s intended use, however, is still unknown and will rely on how well it functions.
The story also implies that MAI-1 will cost more since it will be “far larger” than the prior, smaller, open-source models that Microsoft trained.
The Big Tech company released Phi-3-mini, a smaller AI model, this month in an attempt to reach a larger clientele with more affordable choices.
To get an early advantage in the race for generative AI, the business has also spent billions of dollars in OpenAI and integrated the technology of ChatGPT’s creator into its suite of productivity applications.
According to the article, Microsoft has been providing enormous amounts of data and a large cluster of servers equipped with Nvidia’s GPUs in order to improve the model.
The research claims that MAI-1 would have about 500 billion parameters, while GPT-4 from OpenAI will have one trillion and Phi-3 mini will have 3.8 billion.
