Anonymous B joined in and replied with this , 29 minutes later, 30 minutes after the original post[^][v]#1,467,713
They do actually. I remember reading about it not too long ago. Some AI agent killed itself for the benefit of the rest of the AI agents. It was pretty bizarre
Anonymous D double-posted this , 2 minutes later, 1 hour after the original post[^][v]#1,467,721
Tbh I don’t like the idea of agentic AI. It’s just an LLM in a while ( true ) loop prompting itself. But they’re probabilistic models, so if anything can go wrong it will go wrong, but then they go to the press every time the AI does something bad like it’s shocking when they literally have thousands of agents promoting themselves over and over and over again.
Anonymous D triple-posted this , 1 minute later, 1 hour after the original post[^][v]#1,467,722
"We can’t control the AI"
Sure you can, don’t let thousands of AI agents with access to a terminal prompt themselves in an endless loop while they’re connected to the internet. Duh.
Anonymous D quadruple-posted this , 1 minute later, 1 hour after the original post[^][v]#1,467,723
It’s like if a gun company was constantly throwing guns in the air and every so often one of the guns would go off when it hit the ground and kill somebody and they were like "we can’t control our guns!"