Geoffrey Hinton, widely regarded as the godfather of artificial intelligence, has resigned from Google, warning of the growing dangers posed by AI developments. The 75-year-old British-Canadian cognitive psychologist told the BBC that some AI chatbot risks were 'quite scary' and that he now regretted his life's work.
Dr Hinton's pioneering research on neural networks and deep learning laid the groundwork for systems like ChatGPT. He said chatbots could soon surpass human intelligence: 'Right now, they're not more intelligent than us, as far as I can tell. But I think they soon may be.' He noted that GPT-4 already eclipses humans in general knowledge and performs simple reasoning.
Dr Hinton warned of 'bad actors' using AI for harmful purposes, citing a 'nightmare scenario' where leaders like Vladimir Putin might give robots the ability to create sub-goals, such as seeking more power. He emphasised that digital intelligence differs fundamentally from biological systems, as digital copies can share knowledge instantly, giving chatbots vast knowledge.
Matt Clifford, chair of the UK's Advanced Research and Invention Agency, said Dr Hinton's departure 'underlines the rate at which AI capabilities are accelerating' and called for urgent investment in AI safety. Dr Hinton joins a growing number of experts, including Elon Musk and Yoshua Bengio, who have called for a pause in advanced AI development to implement safety measures.
Despite his concerns, Dr Hinton believes AI will deliver more benefits than risks in the short term and does not advocate halting development entirely. He noted that international competition makes a pause difficult, as China would gain a lead if the US stopped. He stressed that governments must ensure AI is developed with safeguards to prevent it 'going rogue'.
Dr Hinton praised Google as 'very responsible' and said his criticisms carry more weight now that he has left the company. Google's chief scientist responded, acknowledging the need for responsible AI development.
