Well. What does thinking require? Can you answer that question. So far we know that LLMs can handle sequential structure. There have also been papers written th

Well. What does thinking require? Can you answer that question. So far we know that LLMs can handle sequential structure. There have also been papers written that demonstrate that they do build an internal model of the world. Artificial Intelligence is artificial. We humans tend to have a very strong tendency to anthropomorphize everything that moves or has the right shape. That tree stump is not a bear, even when in the dark it may seem so. AI is an entirely different thing from a living being. It doesn't exist on sugars extracted from plants or from the flesh of other animals. Here, the semantics of what "think" means, should be examined more closely. My layman understanding of the brain is that it's not a single thing, it's multiple things working in concert. The "us" is just a figment of our own imagination, built to create a social actor that maximizes the survival of its own genes. We are the fruit of our own genes. So, what "think" means is deeply embedded in human social context, and probably doesn't have much explanatory value when it comes to AIs. Can AIs become dangerous? Obviously, when given goals and understanding of how to grow themselves more numerous and powerful either through manipulating humans or taking electricity and hardware by force. If there ever is a evolution like dynamic, the AI will always grow in a gradient pointing towards more replication. Yeah, maybe I miss the point in your post and the meat of it is that humans are more important than machines, which I do agree with. It's just not an immutable law of the universe, but just something we humans decide.