Speaking on the Today program on Thursday, he said that if firms continue to create AIs that set their own goals, make money and own assets, they will “essentially create a new kind of silicon that will undoubtedly compete with us for resources, no matter how much it cares about humanity and loves us.”
The AI leader’s comments to the BBC followed an essay he published earlier this week in which he warned about Anthropic’s approach to training its AI model Claude.
He sharply criticized Anthropic for training its AI to have human-like qualities, a practice known as anthropomorphization, which he said made it seem as if Claude had his own desires, values and sense of self-worth. The BBC has contacted Anthropic for comment.
In his essay, he warned that tech companies risk creating something that is “impossible” to control if they treat technology like humans.
“AI is not conscious,” he wrote. “They don’t feel, they don’t experience, they don’t suffer. They have no innate preferences or underlying motivations.
“They are sequence completion machines, hollow inside, designed to follow instructions and achieve goals set by people.”
Speaking to the BBC about the upcoming artificial intelligence summit, he called for “alignment” to create technologies that should be “subservient” to humanity.
“I think the good news is that every person will have a very strong interest in making sure that the systems that we all create and… use around the world in every country are safe, controlled and subject to humanity.
“Everything needs to be consistent,” he said.
In his lengthy essay, Suleiman also praised Anthropic boss Dario Amodei and his team for being “thoughtful, principled and intellectually honest people” but nonetheless questioned the company.
Suleiman argued that “consciousness is biological,” stating that “there is no evidence that AI is conscious.”
Suleiman not only called for discussion on the issue, but also said that greater transparency is needed regarding how artificial intelligence systems are trained and evaluated.
This includes independent study of AI behavior and stronger tools to monitor and control the technology, he said.
“We should not, as if in a dream, make a decision that we will later bitterly regret,” he wrote.
Dame Wendy Hall, professor of computer science at the University of Southampton, described the comments as “a conversation we need to have internationally”, contrasting them with “theatrical statements” from some artificial intelligence companies, which she said only served to “scare everyone”.
