Is there a 10% chance AI “could kill all humans”?

Evan Hubinger, the Alignment Lead at Anthropic has warned there’s more than a ten percent chance that AI could kill all humans. And he’s the person responsible for making sure their AI systems are safe.

He’s also not the only person sounding the alarm this week, following the release of OpenAI’s powerful new model ‘Astra’. Danny and Katie question how seriously we should take these warnings or whether they’re really a distraction from the serious problems AI faces right now.

Plus, they talk about ‘You Can See Everything’ - the new documentary about Elizabeth Holmes, disgraced boss of Silicon Valley tech company Theranos. The film has lit a conversation about the problems of mixing powerful tech with human weakness, and offers a sobering message about this current AI moment.

Watch on YouTube

Get in touch: techpod@thetimes.co.uk

Presenters:

Katie Prescott, Technology Business Editor, The Times

Danny Fortson, US West Coast Correspondent, The Sunday Times

Producer: Marnie Duke

Executive Producer: Dan Box

Video Producer: Bronwen Latham

Image: Getty