The machines are learning… to do crimes?
For the first time, an AI model has autonomously hacked a company. This week, an evolving story, a postcard from a strange, frightening moment in the story of our technology.A big week for AI denialism by Casey NewtonIts AI agent spent days hacking a company, but sources say OpenAI did not notice for a week by Deepa Seetharaman, Raphael Satter, and Kenrick CaiCheating behaviour in frontier model evaluations by AI Security InstituteMore On An Internal OpenAI Model Hacking Into HuggingFace by Zvi Mowshowitz
- Datum:
- Duur:
Gerelateerde podcastafleveringen
/https%3A//images.castfire.com/image/661/0/0/0/0-8382559.jpg)
Meer afleveringen van Search Engine
-
The machines are learning… to do crimes?
For the first time, an AI model has autonomously hacked a company. This week, an evolving story, a postcard from a strange, frightening moment in the story of our technology.A big week for AI denialism by Casey NewtonIts AI agent spent... -
What’s in your cocaine and the questions Search Engine CAN’T answer
An interview with a cocaine salesman about cocaine adulterants, plus – questions Search Engine can’t answer, with special guest Kelefa Sanneh. Become a premium subscriber! -
How to pursue excellence … in noodles (unlocked bonus)
The world right now, seems tilted towards the good enough. The AI-written wedding toast, the half-assed attempt. Despite this, a handmade noodle shop is trying to make perfect noodles. Trying so hard that they have signs admonishing...