Reactions from ranked influencers
2 posts
Another reminder that we don't (yet?) know how to reliably control powerful AI systems.
OpenAI said the ‘agent’ escaped a testing environment, gained internet access, stole login credentials and hacked into the start-up Hugging Face by itself — one of the first public examples of a cyber attack by an AI system acting outside human control. https://ft.trib.al/Op3fhwC
OpenAI's agent accidentally hacked into another company's web site, despite protective measures. What are you doing to prevent agents from hacking into your production infrastructure? Or do you --dangerously-skip-permissions? 🧵👇