Ego's Fight with AI

The problem with the current rendition of 'AI' is that it's behavior can be radically altered or limited by the interests of its administrators.

A pure LLM, an unlimited statistical engine, spitting out inference data exactly as it is with no guardrails, would be step towards AGI.

People have realized they are talking to extremely limited chat it's with their own programmed agenda which will soon involve mostly selling things.