Consciousness, intelligence, and all these other properties can be and are mutually exclusive. What will be most useful for humans is a general intelligence tha

Consciousness, intelligence, and all these other properties can be and are mutually exclusive. What will be most useful for humans is a general intelligence that has no motivation for survival and no emotions and only cares about the goals of the human that is in control of it. I have not seen a convincing argument that a useful general intelligence must have goals that evolve beyond what the human gives it and must be conscious. What I have seen are assertions without evidence, "AI must be this way" but I'm not convinced. I can conceive of an LLM enhanced using other ML techniques that is capable of logical and spatial reasoning that is not conscious and I don't see why this would be impossible.