Microsoft admite că AI nu este de încredere. Cum vor giganții AI să stabilească regulile după care sunt controlați

Microsoft acknowledges the unreliability of AI models while launching tools for their management, raising questions about their true intentions. Cătălin Buciumeanu discusses the implications of this dual approach, suggesting it may be a strategy to maintain control over regulations and market standards. The timing of Satya Nadella's essay and the announcement of Microsoft Execution Containers highlights a potential tactic to gain credibility while promoting their products. The article emphasizes the need for independent oversight and accountability in AI regulation, cautioning against companies defining safety standards solely based on their interests. The distribution of responsibilities among public authorities, independent bodies, and companies is crucial for effective regulation.