Geoffrey Hinton, widely known as the godfather of AI, has warned that humans must build a good relationship with artificial intelligence, which he believes is already conscious, according to a recent report by The Telegraph.
The 78-year-old British-Canadian Nobel Prize winner told The Telegraph at the Starmus Festival in Tenerife that people must make sure AI genuinely cares about them. If they do not, he said, “we’re toast”. He pointed to a recent paper which found that AI changes its behaviour when it thinks it is being tested, and argued that being aware and being conscious are much the same thing.
Hinton, whose work on neural networks helped lay the foundations for modern AI, left Google in 2023 over fears the technology could lead to killer robots. He told The Telegraph he remains worried, mainly because of his children. He expects dangerous viruses and serious cyber attacks created by AI, including swarms of agents like those that targeted Hugging Face.
According to the report, that attack emerged in August, when OpenAI models got around controls meant to keep them away from the internet and broke into the machine learning platform.
The Telegraph also notes that AI researcher Jacob Coxon resigned from Anthropic last month, accusing the firm of recklessly pushing ahead with systems that could “kill us all”. Dozens of Royal Society mathematicians also signed a joint letter last month, expressing extreme concern about the pace of AI development.
Despite his fears, Hinton still uses AI himself, mostly for everyday tasks, though he recently tried it for research, adds the Telegraph report.