Microsoft CEO Satya Nadella has begun to take AGI seriously from a political economy perspective. His focus is not on whether current frontier models possess consciousness, but on how enterprises, labor, capital, and democratic institutions will be reorganized as artificial general intelligence becomes widely integrated into economic activities.
During a roundtable discussion following his keynote speech at Microsoft Ignite held at the Chase Center in San Francisco, Nadella stated that he does not believe models like Claude possess consciousness. He also mentioned that Microsoft's AI lead, Mustafa Suleyman, argues against granting models rights or moral status. However, Nadella believes this issue remains worthy of discussion and did not offer an absolute conclusion.
Regarding market structure, Nadella warned that if a single model achieves monopoly status, the impact will extend beyond market cap competition among tech companies. The same model could penetrate sectors such as healthcare, financial services, and even government, thereby amplifying the risks of centralization.
He further suggested that if intelligence exhibits network effects, enabling models to see and learn Everythinggj while human capital is compressed into "token capital," the role of traditional companies could be disrupted. In a scenario with capital but lacking labor, democratic politics would also struggle to function properly.
Nadella thus emphasized the market's need for diverse models, enterprise-owned data, and varied deployment methods. This discussion aligns with Microsoft's direction of promoting enterprise AI tools at Ignite, focusing on the policy and economic risks associated with AI centralization, rather than announcing new products or claiming that AGI has already been achieved.

