Microsoft is reportedly developing a new artificial intelligence (AI) model separate from OpenAI’s widely known GPT. The new model, called MAI-1, is under the leadership of Mustafa Suleyman, a former Google DeepMind co-founder and AI start-up Inflection CEO. With an impressive 500 billion parameters, the MAI-1 model is set to be larger and more expensive than Microsoft’s previous open-source AI models.
The exact purpose of the MAI-1 model is yet to be determined, with its performance being a key factor in shaping its future applications. Microsoft might give a sneak peek of the new AI model at its upcoming Build developer conference. Despite declining to comment on the news, Microsoft’s investment in this new project signals a move towards competing with the likes of Google and OpenAI in the AI space.
The development of MAI-1 follows Microsoft’s recent release of a smaller AI model named Phi-3-mini, aimed at providing cost-effective AI solutions to a broader audience. By investing in OpenAI and leveraging ChatGPT technology in its productivity software, Microsoft has established itself as a frontrunner in generative AI technology.
To support the training and enhancement of the MAI-1 model, Microsoft has dedicated significant resources, including clusters of servers with Nvidia GPUs and vast amounts of data. With an estimated 500 billion parameters, the MAI-1 model is poised to rival OpenAI’s GPT-4 with one trillion parameters, and Microsoft’s Phi-3 mini with 3.8 billion parameters.
Having appointed Suleyman to lead its consumer AI unit and onboarded former Inflection employees, Microsoft is poised for further advancements in the AI domain. While MAI-1 is not directly inherited from Inflection, it may leverage training data from the startup, indicating a strategic approach to AI development.
As Microsoft continues to innovate in AI technology, the unveiling of the MAI-1 model showcases the company’s commitment to pushing boundaries and staying competitive in the fast-evolving AI landscape.
