Microsoft CEO backs slowing AI, but has a 'wake-up message' for Altman, Amodei

Microsoft CEO Satya Nadella has commented on the discussion surrounding the need to slow down the advancement of artificial intelligence (AI), while also delivering a warning to industry leaders such as Dario Amodei and Sam Altman. In a post on the social media platform X, Nadella emphasized that the pursuit of superintelligence must be guided by the fundamental principle that AI should serve humanity and remain under human control.

Nadella went on to stress that the benefits of AI should be accelerated and distributed broadly across nations, communities, and companies. To achieve this, he advocated for the creation of a frontier ecosystem that allows both closed and open-source models to coexist, preventing innovation from being concentrated in the hands of a few entities.

Nadella highlighted the importance of organizations maintaining control over their unique knowledge, stating that every entity should have the capacity to establish its own continuous learning loop without relying on any single model provider.

The Microsoft CEO welcomed research and a deliberate pace in order to ensure that AI alignment is achieved as a design goal. He also endorsed concepts such as "embedded evaluators" and broader mechanisms to ensure AI safety, moving beyond mere rhetoric. Nadella cautioned against the concentration of AI control in the hands of only a few entities, advocating for widespread representation across countries, academia, and industries.

Microsoft's strategy revolves around providing broad access and choice at every point in the AI stack, including enterprise control over learning loops and models. Furthermore, the company plans to publish a "Code of Conduct" for its MAI models, subject to public consultation. Nadella concluded by underscoring that this approach should be adopted by the entire ecosystem, emphasizing the need for broad representation, collaboration, and shared responsibility.