Microsoft next big AI goal building humanist superintelligence for the future

Microsoft has officially launched a specialized research division, the MAI Superintelligence Team, dedicated to pioneering the next generation of advanced artificial intelligence. Led by Mustafa Suleyman, the head of Microsoft AI, this initiative represents a significant financial and strategic commitment to developing what the company calls "practical, grounded, and controllable" technology.

The Shift to Humanist Superintelligence

In the original report, "Microsoft’s next big AI bet: building a ‘humanist superintelligence’", Suleyman emphasizes a departure from the "infinitely capable generalist" models sought by competitors. Instead, Microsoft aims to build a "humanist superintelligence"—specialized systems designed specifically to serve human interests and solve real-world problems.

Key focus areas for the new team include:

• Healthcare: Developing expert-level diagnostics capable of early disease detection within the next 2-3 years.
• Renewable Energy: Designing advanced molecules and improving battery storage technology.
• Education: Creating useful AI companions to assist in personalized learning environments.

Diversifying Beyond OpenAI

While Microsoft remains a major stakeholder in OpenAI, this new project signals a move to diversify its AI portfolio. By recruiting top-tier talent like Chief Scientist Karen Simonyan and experimenting with models from Anthropic and Google, Microsoft is positioning itself to lead the AI race through internal innovation rather than relying solely on external partnerships.

Safe and Grounded Innovation

Addressing the existential concerns surrounding AI, Suleyman clarifies that Microsoft is not building "superintelligence at any cost." The goal is to achieve superhuman performance in specific tasks—similar to DeepMind’s AlphaFold—ensuring the technology remains under human control and delivers measurable clinical and operational benefits.