How Did Microsoft Fumble the AI Ball So Badly? (schneidenba.ch)

Microsoft's ambitious Copilot suite has faced significant setbacks, with disappointing adoption rates among its enterprise customers. Despite a massive ecosystem of 440 million Microsoft 365 subscribers, Copilot has only attracted approximately 8 million active users, yielding a mere 1.81% conversion rate over nearly two years. Critics cite the product’s lack of perceived value—ex-employees and consultants question whether the $30/month charge delivers sufficient return on investment. Notably, many enterprises like Amgen have reverted to using ChatGPT instead, highlighting Copilot's struggles in delivering a superior user experience. The significance of these developments extends beyond mere market performance; Microsoft’s reliance on OpenAI’s models has hampered their ability to innovate and adapt at speed. Unlike Google’s Gemini, which benefits from in-house development and control, Microsoft is left with limited influence over the improvements and roadmap of its Copilot products. As Google continues to enhance its AI offerings, Microsoft risks falling further behind, raising questions about its competitive strategy and the long-term viability of products that rely heavily on external technology partnerships. This situation underscores the challenges of navigating dependencies in the rapidly evolving AI landscape.