AsianFin--Microsoft is training a new, in-house AI language model to match the scale and sophistication of AI models from tech giants Google and OpenAI, The Information reported on Monday.
This marks the first time Microsoft has built its own AI model of this magnitude since investing over $10 billion in OpenAI for the rights to reuse the startup's AI models. OpenAI's GPT-4 powers not only ChatGPT but also Microsoft Copilot.
The new model, internally referred to as MAI-1, is being overseen by recently hired Mustafa Suleyman, the Google DeepMind co-founder and former CEO of AI startup Inflection, the report said, citing two Microsoft employees with knowledge of the matter.
According to The Information, Suleyman, previously associated with Google DeepMind and the CEO of AI startup Inflection, brings extensive expertise to the project. While Microsoft acquired Inflection and its intellectual property for $650 million, MAI-1 is an independent initiative.
The scale of MAI-1 is noteworthy. The model is expected to feature approximately 500 billion parameters, significantly larger than Microsoft's previous AI offerings, The Information reports. Such complexity demands substantial computing power and resources for training.
The advancement of MAI-1 indicates a dual strategy for AI within Microsoft, emphasizing both small-scale, locally operated language models for mobile devices and larger, cutting-edge models driven by cloud computing.
According to a source from The Information, the precise purpose of MAI-1 has yet to be defined, even within Microsoft. Its optimal utilization will be contingent upon its performance. To train the model, Microsoft has dedicated a significant cluster of servers equipped with Nvidia GPUs and has compiled training data from diverse sources, including text generated by OpenAI's GPT-4 and publicly available Internet data.
Depending on the progress made in the coming weeks, The Information reports that Microsoft may preview MAI-1 as early as its Build developer conference later this month.
Microsoft's commitment to AI is reflected in its increased capital expenditures. The company's spending surged by 79% to $14 billion, with a significant portion allocated to AI infrastructure, according to CNBC.
However, Microsoft faces challenges in meeting the burgeoning demand for AI computing power, with a shortage of data center infrastructure impacting deployment capabilities.
AI services are essential components of Microsoft's Azure cloud computing platform. Azure's revenue saw a significant 31% rise, with AI playing a crucial role in driving this growth. However, supply limitations, such as those from hardware suppliers like Nvidia, present challenges to Microsoft's aspirations in the AI sector, CNBC reported.
The rise of generative AI models has fueled demand for AI-powered features across various products and services. Microsoft has responded by integrating AI assistants into its offerings, such as the Teams communication app and the Bing search engine.
Microsoft plans to intensify its investment in cloud and AI infrastructure. The company aims to increase capital expenditures, particularly for cloud infrastructure, to address escalating demand. Azure's AI services are a critical battleground in Microsoft's rivalry with Amazon Web Services.