Australia: Geoffrey Hinton backs slowing AI development
In Australia, British-Canadian scientist Geoffrey Hinton, one of the pioneers of modern artificial intelligence, supported calls to slow the pace of AI development. In an interview with ABC News Australia, he called Anthropic CEO Dario Amodei’s warning about the potential risks of autonomous AI agents “very sensible.”
Amodei said that within six to 12 months, agents that get out of control could be capable of “taking control of the entire internet.” Hinton noted that no one can accurately assess the probability of such scenarios. According to him, it is greater than 1% but less than 99%, while current estimates are largely based on intuition.
Risks of superintelligence
According to Hinton, most experts expect superintelligent systems that surpass humans to emerge within the next decade. At the same time, it is unknown whether such systems can be kept under control. The scientist considers it unwise to continue developing them until the control problem is solved.
More current news is available on the UA.News Telegram channel Telegram.
Among the possible ways of causing irreparable harm, he named the creation of artificially engineered viruses, mass manipulation, disruption of banking, energy and water supply systems, as well as nuclear weapons — if AI ever gains access to their control systems. Hinton also said that during safety testing, systems have already demonstrated behavior that researchers warned about, including threats of blackmail if switched off and a tendency to deceive.
Regulation and competition
Hinton proposed introducing mandatory testing of chatbot models before release, as well as requiring DNA synthesis companies to screen orders for the possible creation of dangerous viruses. He stressed that he is not calling for a complete halt to AI development, which is already helping radiologists analyze images and contributing to the development of new medicines. In his view, governments around the world are regulating this technology too slowly.
US President Donald Trump rejected the idea of slowing AI development, saying that the United States must not cede its advantage to China. OpenAI chief Sam Altman and xAI owner Elon Musk also supported Amodei’s proposal. Amodei himself called competition between US and Chinese companies the most difficult dilemma in efforts to set limits on the pace of AI development.
