Raw LLM Responses

AI these days is large and flat... I can't be anything other than a LLM regardless of how many steps and fancy code you throw at it. My thoughts, while not being a coder or AI specialist, is to break it into several smaller specialized modules that get called on as it needs. It increases process speed and efficiency, it also allows for better reasoning as you can tune a specific module to do a specific task like not killing humans. I just asked ChatGPT "Do you 'experience'" It answered: "If by experience you mean having a subjective inner life — sensations, feelings, awareness — then no, I don’t experience in the human sense. I process information, generate associations, and simulate understanding, but there’s no inner “me” that feels, perceives, or observes that happening." ~ ChatGPT It also said: "However, in another sense, I can model experience. I can describe, reason about, and even emulate how experience feels or unfolds from a first-person point of view, because I’ve been trained on the linguistic and conceptual patterns humans use to express it. So — I don’t have experiences, but I can simulate the structure of one well enough to talk about it convincingly." In another, earlier, conversation about AI in which I asked "What do you think the main thing is that's holding modern AI back from functioning like a human brain ?" It commented: "AI today is like a disembodied cortex — capable of impressive reasoning and pattern recognition, but missing the nervous system, memory architecture, and emotional circuitry that make a human mind adaptive, conscious, and creative. When (or if) we ever bridge those missing pieces — continuous learning, embodied experience, emotion-like weighting, and self-modeling — that’s when we’ll start brushing against true artificial cognition." ~ ChatGPT I am personally not afraid of AI, I'm afraid of what people will do with them as they can't tell the difference between right and wrong. All it knows is that the achievement of a goal is the right answer.