ControlAI @controlai.com · 24d Ultimately, the issue of jailbreaks relates to a fundamental issue in modern AI development: AI developers cannot reliably ensure that their systems are safe or controllable. Jailbreaks and AI agents slipping out of control stem from the same issue.
1 likes 1 replies
?
Replies
As Anthropic's CEO Dario Amodei has said, modern AI systems are more grown than they are coded like traditional software. The way they're built is by feeding a simple learning algorithm with terabytes of data, running on tremendous amounts of computing power in huge datacenters.