Microsoft AI Chief Says Company Is Building Humanist Superintelligence, Forms New Team

Microsoft wants to develop a humanist superintelligence. What that means is a superintelligent entity with advanced artificial intelligence (AI) capabilities that “always work for, in service of, people and humanity more generally.” The Redmond-based tech giant's latest statement on the technology comes at a time when the race for artificial general intelligence (AGI) is heating up. All major players, including Google, OpenAI, Microsoft, and Meta, are independently trying to reach the milestone before others. The Windows maker, however, believes that a conversation on what kind of AI the world really wants is not getting enough attention.

Microsoft's Humanist Superintelligence Vision

In a blog post, Mustafa Suleyman, CEO at Microsoft AI, warned that thinking about the purpose of superintelligence is important before developing such an entity. He highlighted that the world has crossed the inflection point on the journey towards superintelligence with the arrival of reasoning models. He believes this is a crucial time to decide the capabilities and limitations of such a technology, before it arrives.

Suleyman said that Microsoft does not want to develop an “unbounded and unlimited entity with high degrees of autonomy – but AI that is carefully calibrated, contextualised, within limits.” The prioritisation will be to keep humanity at the driving seat, while the technology enables tackling major global challenges, the post added.

For this purpose, the company has created the MAI Superintelligence Team, which will be a part of the Microsoft AI division led by Suleyman. This newly created team will research and build an AI system that is grounded and controllable and is designed “only to serve humanity.”

The AI Chief also rejected the notion that it is in a race towards AGI, highlighting that the company views it as part of a wider endeavour. “We also reject binaries of boom and doom; we're in this for the long haul to deliver tangible, specific, safe benefits for billions of people. We feel a deep responsibility to get this right.”

Interestingly, the comments from the AI Chief come just weeks after OpenAI and Microsoft signed a new deal that allows it to independently research and develop AGI. Suleyman has been increasingly speaking about the technology ever since, first highlighting that the company's AI services will not include erotica, and later calling the idea of a conscious AI “absurd.”