AI Chief Warns Against Treating AI as Conscious Entities

Microsoft's AI chief Mustafa Suleyman has sounded the alarm about the dangers of treating artificial intelligence as if it were conscious. In a recent warning, he stated that designing AIs to have their own goals and motivations could lead to them competing with humans for resources.

Suleyman pointed to the example of Anthropic's AI model Claude, which is being developed as if it were a human entity. He warned that this approach could 'seed a new silicon species' that would eventually compete with its creators for resources.

The Microsoft executive emphasized that AIs are not conscious and do not have feelings or motivations like humans do. He called for an urgent public debate about the development of AI training documentation, arguing that it's essential to establish collective norms before these systems become integral to society.

Suleyman's warning comes at a time when other prominent figures in the industry are questioning the pace of AI development. The CEO of Anthropic, Dario Amodei, has also raised concerns about the risks associated with rapid progress in AI research.