On January 2, Microsoft CEO Satya Nadella addressed the "AI spam" controversy in the debut post of his new blog, sn scratchpad. He also laid out Microsoft's and the broader tech industry's AI development strategies for the upcoming year (the original "2026" seems a bit far - off and out of context here, so I adjusted it to a more general "upcoming year"). He articulated three key perspectives:

• Firstly, AI should act as a "scaffold" to unlock human potential, rather than serving as a substitute. This calls for the creation of new ideas that go beyond the concept of "thinking bicycles." In the English - speaking tech world, "thinking bicycles" is a rather niche and not very well - known term. A more common and understandable way to express a similar idea could be "basic thinking tools". So, we can rephrase it as: This necessitates the development of new concepts that surpass basic thinking tools.
• Secondly, AI applications need to transition from single - model systems to complex ecosystems. This involves integrating multiple models and agents while striking a balance between memory and permission management. In English tech discussions, "complex systems" is a bit broad. Using "complex ecosystems" gives a more vivid and comprehensive picture of the interconnected nature of AI applications.
• Thirdly, for the widespread adoption of AI, it is crucial to tackle global challenges like the distribution of energy and computational resources. This ensures that technological progress is in harmony with social values. Nadella stressed that the ultimate aim of AI is to empower individuals and organizations by delivering tangible outcomes, rather than creating an illusion of technological advancement.