Yeah, setting up fake identities and recruiting unwitting human accomplices via social manipulation is a pretty key part of take-over schemes. This is something language models are already close to competent enough linguistically for (although not agentic / goal-driven to do so, thankfully).
Yeah, setting up fake identities and recruiting unwitting human accomplices via social manipulation is a pretty key part of take-over schemes. This is something language models are already close to competent enough linguistically for (although not agentic / goal-driven to do so, thankfully).
I mean, they don’t have to be unwitting; Blake Lemoine knew he was working with a computer.