Microsoft has officially joined the AI cost-cutting trend by increasingly relying on its own in-house models. According to recent reports from TechCrunch and The Decoder, the tech giant is gradually replacing expensive models from external partners like OpenAI and Anthropic with its proprietary MAI models for popular office applications.
Diễn biến chi tiết
According to reports from The Decoder, Microsoft has already begun deploying its in-house models, named MAI, into core products within the Microsoft 365 ecosystem, such as Excel and Outlook. The system already processes tens of thousands of queries per week through these proprietary models. This represents the first concrete step in a long-term plan to reduce financial dependency on the AI startups in which the company has invested billions of dollars. Microsoft's AI chief, Mustafa Suleyman, expressed his ambition to "ultimately eliminate" the costs of external models.
Phân tích kỹ thuật & Công nghệ
Shifting from leading commercial large language models (LLMs) such as OpenAI's GPT-4 or Anthropic's Claude to the proprietary MAI model family marks an architectural pivot in Microsoft's software engineering. While specific technical details of the MAI series have not been widely disclosed, analysts suggest these models are specifically optimized for targeted office tasks to minimize the required computational resources on Azure cloud infrastructure. However, this raises questions regarding the maintenance of accuracy and versatility compared to previous large-scale, general-purpose models.
Ý kiến chuyên gia & Nhận định
According to industry experts, this strategy could be a double-edged sword for end-users. The Decoder points out that for Copilot customers, this transition could mean experiencing less performance for the same subscription price. This shift highlights the severe financial pressure on tech giants as AI infrastructure operational costs expand beyond initial margin projections.
Tác động & Tương lai
Microsoft's cost-cutting trend reflects a new reality in the global tech market, where the phase of pure feature competition is gradually giving way to economic efficiency. Achieving autonomy in foundation models not only helps Microsoft optimize margins for its cloud services but also reshapes its complex frenemy relationship with partners like OpenAI in the near future. For global users, this change could directly impact the response quality of daily AI assistant tools.