Microsoft raises Microsoft 365 pricing to drive AI computing toward user devices to reduce costs

With the rapid development of large-scale generative artificial intelligence technology, Microsoft has recently taken a series of compelling measures in its products and services to cover huge operating costs. The company not only increased the subscription price of its Microsoft 365 software, up to 45%, but also launched product versions with advertising and canceled some data center rental plans.

Microsoft CEO Satya Nadella recently said that despite the company's huge investment in AI technologies, the value brought by these technologies is not significant so far. This series of measures seems to reflect that Microsoft is not retreating in AI investment, but is exploring new profit models to try to pass costs on to consumers.

The cost of generative AI is quite expensive. Take OpenAI, a market leader, for example, although the company achieved $3.7 billion in revenue last year, spending as much as $9 billion, causing a net loss of about $5 billion. The operating cost of OpenAI mainly consists of two parts: model training and inference. As the user base grows, the cost of reasoning is also increasing.

As the largest investor in OpenAI, Microsoft is also undertaking this series of cost pressures. Therefore, Microsoft is seeking to reduce these expenses by shifting more AI computing tasks to users’ devices. Users need to bear the hardware and its operating costs themselves, which may reduce Microsoft's operating burden.

In addition to raising prices, Microsoft has added a dedicated C opilot button to its new devices, encouraging users to perform AI processing on their own devices. This "edge computing" strategy aims to reduce energy consumption and resource waste in data centers, while also improving user privacy protection.

However, this strategy also presents some challenges. As AI computing costs are passed on, consumers may need to update devices frequently, which will generate more electronic waste. Furthermore, hardware differences may lead to uneven user experience, especially in the field of education, which may exacerbate inequality.