Microsoft is launching a new team focused on advanced artificial intelligence research, called the MAI Superintelligence Team, led by Mustafa Suleyman. Suleyman, who previously co-founded DeepMind and led AI startup Inflection, aims to develop practical AI technology designed to address real-world problems while ensuring it remains controllable. He emphasizes that the goal isn't to create an abstract superintelligence but to serve humanity through useful applications. The formation of this team comes in the wake of Meta's investment in its own AI initiatives. Speculation around superintelligence often involves machines surpassing human intelligence, but Suleyman's approach prioritizes grounded applications. Microsoft has been expanding its AI capabilities rapidly since the ChatGPT launch in 2022, incorporating OpenAI’s models into its Bing and Copilot products. The company also holds a $135 billion equity stake in OpenAI, following a major restructuring. To diversify its AI resources, Microsoft has begun leveraging models from Google and Anthropic, aiming to reduce reliance on OpenAI. Suleyman’s team plans to develop AI companions that enhance education and explore applications in medicine and renewable energy. He highlights the importance of achieving expert-level performance in diagnostics and operational planning within clinical settings. Amid concerns over excessive spending in the AI sector without clear profit strategies, Suleyman reassures that their approach won't compromise safety or ethics for advancement.