FUTURATI PODCAST cover image

Ep. 129: Applying the 'security mindset' to AI and x-risk | Jeffrey Ladish

FUTURATI PODCAST

CHAPTER

The Risks of Having a GPT Three Level Model

I'm basically saying that like it's really risky to have one without the other. So you touch a GPT three level model and you gave it this engine it actually probably becomes an interestingly capable agent, but not not posing much risk of catastrophe at that moment. But currently the way we train it or whatever it just doesn't have that engine. And if you ask me to make a detailed prediction like that would be a nonsensical way of approaching that. I often see people pick apart the specific scenarios that somebody like Ellie as your proposes, but nobody's saying that's how it will happen. It's just a scenario that occurred to a human level intelligence that is slightly to the right on the

00:00
Transcript
Play full episode

Remember Everything You Learn from Podcasts

Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
App store bannerPlay store banner