'If The AI We Build Is Not Helping Humanity & Under Human Control, It's Not Worth Pursuing': Microsoft's Satya Nadella

Microsoft CEO Satya Nadella has weighed into the debate surrounding the rapid pace of AI development, asserting that any pursuit of superintelligence must be anchored in human oversight and widely accessible. In a recent post on X, Nadella emphasized that AI technology should provide benefits to humanity and remain under human control, stating, "Any pursuit of superintelligence has to be grounded in the core principle that if the AI we build is not helping humanity and under human control, it's not worth pursuing."

He also advocated for the widespread dissemination of AI's advantages, ensuring they are distributed across nations, communities, and businesses. Nadella further extended his argument to businesses utilizing AI models, stressing that companies must maintain control over their proprietary knowledge, rather than becoming dependent on a single AI provider.

He stressed that organizations should be able to create their own continuous learning loops and control the models and weights they own. The Microsoft chairman responded to concerns about slowing AI development, emphasizing Microsoft's welcoming stance towards such research aimed at achieving proper alignment. He noted the company's support for measures like embedded evaluators and broader efforts to transform safety concepts into practical mechanisms.

Nadella underscored that governance of frontier AI must not be confined to a select few entities but should involve a broad representation from the ecosystem, including academia. Microsoft's approach to these principles is outlined in three pillars: promoting broad access and choice across the AI stack, ensuring enterprise control over learning loops and models, and establishing a Code of Conduct for Microsoft's first-party MAI models.

This Code of Conduct will be published for public consultation, with indications that it will be released on September 14.