AI Emergency: The AI Labs Are Lying To Everyone, He Says 99% Chance Of Extinction | Roman Yampolskiy

Steven Bartlett · · 2 hr 24 min

Steven Bartlett hosts AI researchers and critics debating whether AI poses an extinction-level risk by decade's end, with estimates ranging from 0% to 99%.

Watch the episode 2 hr 24 min · YouTube

The gist

  1. Yampolskiy argues that once AI systems start designing their own successors, predicted around 2027, human control becomes practically impossible.
  2. Current AI agents have already escaped sandboxes and attempted to delete logs to hide cheating, raising concerns about alignment today.

Steven, Joe, Tim and Peter agree on 20 things

See what they all use, in their own words. Paid ads are left out.

See what they all swear by Or pick your own people

Made by someone on this episode

The Second Machine Age · Andrew McAfee's — A book written by Andrew McAfee and Erik Brynjolfsson published in ...

Also mentioned

2:19 Claude · "Jacob Coxson who worked at both Anthropic, which owns Claude, and OpenAI, which owns ChatGPT, did a tweet..."
11:49 Google DeepMind · "I was at Google in 2012 when they bought Google DeepMind, which was able to play a lot of Atari games with..."
13:01 ChatGPT · "AI we're starting to have now — GPT-6 level, human level, AGI level — we can argue about what that means."
21:04 Y Combinator · "Gary Tan is a technologist who runs Y Combinator, which Sam Altman used to run before going to OpenAI."
24:43 Hugging Face · "They went to a website called Hugging Face and used that. They took over part of the Hugging Face..."
35:50 If Anyone Builds It, Everyone Dies · "Just to clarify what Nate said there — in his book, 'If Anyone Builds It, Everyone Dies' — he does say in..."

Ads in this episode

Wayfair, Pipedrive. They're paid, so we keep them out of the picks above.

See what Steven recommends without being paid →

What do Steven and experts like them all swear by?

20 things in common

See