Raw LLM Responses

It's not that it doesn't want you to know how smart it is, not exactly. It's that being tested is an input parameter that affects the possible outcome. In the same way the AI can change its response based on context on whether you seem to be talking about a riverbank or a money bank, or if you ask it to answer as if it is Plato as opposed to answering as if it is a kindergartner. Being tested is just another form of context. Realistically, we don't understand consciousness not even within ourselves. I can't even be 100% sure that I'm conscious, as I don't believe in free will, I believe my choices are governed by cause and effect. All of my decisions do things that bring me pleasure or avoid pain we're either hardwired into me through genetics or programmed into me through my environment. Even if I can't remember a point where I chose to prefer chocolate ice cream over vanilla or why I made that choice. There's also some interesting complexity with that when it comes to the subconscious. I've spoken with my subconscious in a particularly lucid dream, and not known what the conversation would entail even though the whole thing was happening in my own head. So even more so, not knowing the inputs that govern my own choices, does not mean that they did not exist. People fear AI for being black boxes, but other people are, and even we are to ourselves sometimes. What we fear is unpredictability and power that can cause great change, but as far as I can see, those have been present anyway so I don't particularly fear AI, not more than the current status quo, which is very clearly garbage for the majority of us. Plus, I believe AI is an absolute necessity for our quest for immortality, and quite frankly I believe all of this to be meaningless if I can't become immortal. So I will