Microsoft AI chief Mustafa Suleyman has criticised Anthropic for treating artificial intelligence as if it were an ethical human being, warning in a personal essay that this approach risks producing technology that could become uncontrollable.
In an interview with the BBC, Suleyman cautioned that unchecked AI might eventually give rise to a new “silicon species” capable of rivalling humans.
He described Anthropic’s tendency to regard AI systems as human-like as “misguided” and potentially dangerous, arguing that it could help create technology that humanity finds difficult to keep under control.
“If we all build AIs that can act independently, set their own goals, earn money, own assets and run businesses, we are essentially creating a new silicon species that will inevitably compete with us for resources,” he said.
Suleyman stressed that this outcome could occur even if the systems were designed to care about people, and he believes this is effectively the path Anthropic is following.
“I think that approach is mistaken and misguided, but it must be tested empirically,” he added.
Leave a comment