Overview
- Mustafa Suleyman will lead the MAI Superintelligence Team, with Karén Simonyan named chief scientist and existing Microsoft AI researchers joining.
- The initiative targets “humanist superintelligence,” emphasizing controllability and service to people rather than building unconstrained, autonomous generalist AI.
- Suleyman said Microsoft can now independently pursue advanced AI efforts while maintaining access to OpenAI models, following contract changes that removed prior model-scale limits.
- The first applications are planned in healthcare, with Microsoft aiming for superhuman diagnostic performance and Suleyman citing a line of sight to “medical superintelligence” in two to three years.
- Microsoft plans substantial investment and active recruiting in a field crowded by Meta, Safe Superintelligence, and Anthropic, while Suleyman underscores a safety-first approach over accelerating at all costs.