Show Notes
On 31 March this year, the British scientist Geoffrey Hinton resigned from Google, where he had directed AI research for a decade. Artificial intelligence, he argued, had reached the point where it could rapidly surpass human intelligence, and potentially take control: it was now an existential risk. One of the three ‘godfathers of AI’, Hinton won the Turing Award, the Nobel of computing, in 2018. Now the three scientists who share the award are divided: Yoshua Bengio shares Hinton’s fears and is calling for caution, while Yann LeCun believes AI will bring positive change. In this New Statesman cover story, Harry Lambert visits Hinton at his home in London for a fascinating extended profile of the man at the heart of today’s debate about AI. He talks to Hinton’s critics, who might disagree on the pace of change but agree that further research and oversight are needed. Are we looking at an existential threat, humanity’s salvation, or both?
This article originally appeared in the 24 June issue of the New Statesman. You can read the text version here.
Written and read by Harry Lambert. If you enjoyed listening to this, you might like Margaret Atwood: why I don’t write utopias.
Hosted on Acast. See acast.com/privacy for more information.
More from Audio Long Reads, from the New Statesman
How did one detective take on an international network of romance fraudsters?
They no longer have a stranglehold on Oxbridge and would lose tax breaks under Labour. So what is elite education really selling?
On 2 November 2023, Rishi Sunak closed his global AI Safety Summit at Bletchley Park by interviewing the richest man on Earth, Elon Musk. The mood was deferential (the PM towards the tech billionaire). Was Sunak eyeing up a post-politics job in San Francisco, some wondered, or calculating that Musk’s Twitter might be an effective campaigning tool come 2024?
What might be the long term impact of the Israel-Hamas war on global alliances? In this week’s audio long read, the New Statesman’s contributing writer John Gray reflects on three weeks of bloodshed, beginning with the massacres of 7 October, and their wider consequences. An escalating conflict will empower Iran and Russia, he writes, as well as strengthen swing states such as Saudi Arabia and Qatar. The United States might abandon Ukraine, or dilute its commitment to defending Taiwan from China. And with a presidential election on the horizon, does the White House have the stamina for a protracted foreign war?
In May this year, an American woman sought the help of a chatbot on an eating disorders website. The bot, named Tessa and running on an evolving, generative AI, advised her to start counting calories. Perhaps she should get some calipers, it suggested, to measure her body fat. When it emerged that Tessa had given similarly dangerous advice to others, the bot was taken down.