'AI Godfather' Geoffrey Hinton: "Killer Robots Will Appear Within 10 Years"

Geoffrey Hinton, the 'Godfather of AI' and Professor Emeritus at the University of Toronto, Canada, who first devised the concept of deep learning, the foundation of artificial intelligence (AI) training, predicted that "within 10 years, autonomous robot weapons capable of killing humans will emerge."

'AI Godfather' Geoffrey Hinton, Professor at the University of Toronto [Image provided by Professor Geoffrey Hinton]

In an interview reported by Nihon Keizai Shimbun on the 10th, Professor Hinton warned about the dangers of AI and made this statement.

Professor Hinton said, "If you give AI a goal, it might find a harmful way to humans as a solution. For example, if AI is instructed to stop climate change, there is a concern that it might think excluding humans is necessary to achieve this goal and carry it out."

He also expressed concern that competition between different AIs could lead to a battle over data centers. He predicted, "No one wants a future ruled by AI, and this will form the basis for countries to coordinate regulations on AI weapons."

Professor Hinton stated, "I believe that large language models, which are the foundation of generative AI, understand language just like we do." He added that he uses whether AI understands jokes as a criterion for judgment, mentioning that when he asked a chatbot based on Google's 2022 language model (PaLM) to explain a joke, the chatbot explained why the joke was funny in full detail.

He said, "Among philosophers, there is a claim that humans are innately equipped with the ability to use language, but that is wrong. I believe language is acquired after birth. From the perspective of subjective experience, I think AI can have senses similar to humans."

It is reported that Professor Hinton, who founded the AI company 'DNN Research' while a professor at the University of Toronto, continued his research activities as part of Google after the company was acquired by Google in 2013, but resigned in April last year to raise awareness about the dangers of AI. Since then, he has been living in Canada and advocating the risks of AI.