Why the real AI apocalypse is already here
Summary
Kai Wright and Carter Sherman discuss the surge in bipartisan concern over AI, highlighted by a joint event between Bernie Sanders and Steve Bannon. They speak with Guardian tech reporter Johanna Buya about recent AI agent "hacks" and whether these incidents signal a loss of control. The conversation explores the skepticism surrounding AI executives' calls for self-regulation, noting potential conflicts of interest and PR motives. Buya also points to immediate, real-world harms, including AI use in surveillance, bioweapons development, and geopolitical espionage, arguing that current danger…
Topics discussed
Sponsor: Wayfair Halloween Decor
Intro: AI companies' PR battle and public alarm
Show intro: The 'end of humanity' headlines
Introduction of tech reporter Johanna Buya
Bernie Sanders and Steve Bannon's bipartisan AI rally
Defining 'super intelligent AI' and the new political moment
Historical context: From Hawking's warning to ChatGPT
The 'hacks' as evidence of loss of control
Explaining the OpenAI agents hacking Hugging Face
Why AI is different from traditional rules-based software
The two-week pause and skepticism about company claims
Critique: Is this a strategy to block competitors?
The 'Turkeys voting for Christmas' self-regulation proposal
Getting ahead of public opinion and midterm elections
Light banter about contacting Sam Altman
The PR battle: Balancing genuine fear and marketing
Bipartisan opposition to data centers and surveillance
Current dangers: Climate change and surveillance tech
Anthropic's 160-page disclosure on AI misuse
Specific harms: Bioweapons, surveillance, and cyber hacks
Closing remarks and credits
Outro and teaser for next episode with Naomi Klein
Sponsor: Wayfair Halloween Decor (Repeat)
Listen ad-free on Castria