Microsoft AI CEO admits Gemini 'can do things that Copilot can’t do' (www.techradar.com)

In a candid interview, Microsoft AI CEO Mustafa Suleyman acknowledged that Google’s Gemini 3 can outperform Microsoft's Copilot in certain capabilities, marking a rare admission in the competitive landscape of AI models. While he highlighted Copilot's strong performance in real-time assistance and digital vision—allowing users to share screens and receive instantaneous feedback—Suleyman emphasized a more grounded vision for Copilot as an intuitive personal assistant designed for everyday tasks. This differentiation underscores Microsoft's commitment to utility over raw power in AI development. The discussion signals a shift in the AI/ML community, wherein nuanced comparisons between competing models like Gemini 3 and Copilot may benefit consumers by clarifying their unique advantages. Gemini 3 is positioned as a cutting-edge multimodal model aiming for creative problem-solving and superior understanding of complex queries, while Copilot focuses on practical applications across Microsoft’s suite of products. Suleyman's concept of "humanist superintelligence" reflects a responsible approach to AI development, prioritizing safety and usability without compromising innovation. This dialog around distinct capabilities could lead to more tailored solutions that cater to diverse user needs.