Deep learning pioneer Geoffrey Hinton leaves Google, wants to warn about AI
One of Google’s top AI researchers, Geoffrey Hinton, is leaving the company, partly due to his age of 75. Now he wants to warn about the possible dangers of the AI he helped develop; something he now says he regrets.
Geoffrey Hinton
Hinton resigned from Google in April and announced this news worldwide on Monday, including in an interview with The New York Times and the BBC. He wants to be able to speak out about artificial intelligence and warn about the possible consequences, which he said he could not do as a Google employee. Hinton emphasizes that his departure does not mean that he wants to be critical of Google, a company that, according to the scientist, handles AI ‘extremely carefully’.
At the same time, the researcher fears that the current power struggle between companies such as Google and Microsoft to create the best generative AI could have negative consequences. Hinton currently fears that many fake images, videos and texts created using generative AI tools will be distributed in the near future. The average person would therefore no longer know what is and is not truth, Hinton fears. The British-born is also afraid of the possible major changes in the job market.
Besides the immediate future, Hinton fears long-term consequences, such as autonomous weapons. “Some bought into the idea that artificial intelligence could actually become smarter than humans,” Hinton says. “But most people thought that was far away. I thought so too. I thought it was 30 to 50 years, or maybe even further away. Obviously I don’t think that anymore.” Legislation may not help with this, because such research can be carried out secretly by companies. Hinton thinks it is better if scientists work together to control artificial intelligence.
Geoffrey Hinton is a pioneer in the fields of deep learning and neural networks. Early last decade, together with students from the University of Toronto, he trained neural networks to recognize objects in images and predict letters in sentences. writes MIT Technology Review. These are techniques that form the basis of generative AI, such as the chatbots that have been released since the end of last year. Hinton has worked at Google since 2013 and was a Google Fellow, a title for the most important developers or researchers within the company. The scientist is also known as the ‘Godfather of AI’.