Raw LLM Responses

LLMs dont actually know anything is the problem. It is concerning, but more in the normal, dumb, human way of "lets plug cracked-out cleverbot into government data centers so we can fire 100 employees." The fidelities of these systems are grossly overinflated and these companies only benefit from exaggerating their capabilities. The greatest capability i see from AI is simple mass pattern recognition, because thats all these engines are. Theyre fantastic at predictive uses but theyre godawful at anything genuine. The reason these tech bros focus on the economic aspects of this is because, despite the things they are good at, the marketing of LLMs as AI is like technological Radithor imo