People Are Sharing The “Silent Threats” That Will Become A Huuuge Problem If Not Addressed Sooner

9.

“AI. I’m not kidding either.”

—roppunzel

Some leading AI scientists warn that AI “agents” or “agentic AI,” aka AI systems that operate autonomously without constant human supervision, could prove dangerous if they do not behave in line with human goals. AI pioneer Yoshua Bengio, notably called one of the “godfathers of AI,” has warned that some agentic AI models are already edging towards “dangerous behaviors” traits such as “deception, cheating, lying, hacking, self-preservation, and more generally, goal misalignment.”