Microsoft forms superintelligence team under AI chief Suleyman

On Thursday, Microsoft announced the establishment of a new team dedicated to pioneering advanced artificial intelligence research. Mustafa Suleyman, the head of Microsoft's AI division—which includes the popular Bing and Copilot assistant—shared this development in a blog post, confirming that he will lead the newly formed MAI Superintelligence Team. Suleyman emphasized that the team's mission is to tackle tangible challenges while ensuring the technology remains manageable and grounded. He stated, "We are not building an ill-defined and ethereal superintelligence; we are creating a practical technology specifically designed to benefit humanity." This initiative comes on the heels of significant investments made by other tech giants, particularly after Meta's substantial hiring spree for its Meta Superintelligence Labs unit, which focuses on similar research. The term 'superintelligence' refers to systems that exceed human intelligence, a concept that has gained traction in recent years. Suleyman, a co-founder of the renowned AI lab DeepMind—acquired by Google in 2014—has a rich background in AI. Following his departure from Google in 2022, he co-founded and led the AI startup Inflection, before being recruited by Microsoft along with several colleagues from Inflection. In the race to enhance their offerings with generative AI functionalities, tech companies are rapidly onboarding top AI talent. The surge in interest began with the launch of OpenAI's ChatGPT in 2022, and Microsoft has integrated OpenAI's models into its products like Bing and Copilot, while also maintaining a significant $135 billion equity investment in OpenAI. In a strategic move to diversify its AI capabilities, Microsoft has begun to explore models from Google and Anthropic, a company founded by former OpenAI executives. Suleyman outlined that the focus of the new AI research group will be on developing useful tools for education and other fields, while also delving into specific sectors such as healthcare and renewable energy. "We aim to achieve expert-level performance across a wide range of diagnostics, along with effective planning and prediction capabilities in clinical settings," Suleyman detailed in his blog post. Amid rising concerns from investors about excessive spending on AI without clear profitability, he reassured that the goal is not to pursue superintelligence recklessly or without constraints.