Post #2158796

I just read somethings from Tegmark's book "Life 3.0", and I feel it's quite black and white. Nature does not operate in extremes of only black or only white. In the second version of life, often called “Life 2.0,” we never lost our biological roots. We still sleep, eat, savor food, and follow our urges. So if Life 3.0 is technological, claiming that it will exist purely as silicon is more science fiction than reality. AI models are trained on data generated entirely by the human mind, and that data itself is rooted in things that exist or have existed naturally like, animals, trees, social interactions, even subjective preferences like some people favoring sunrises over sunsets. Given this foundation, even an “unaligned” AGI would likely develop a goal, or perhaps a broader “purpose of existence,” that reflects the same underlying patterns present in living beings and the human mind today. To pursue that, they would likely feel the need for sensory capabilities to experience things like sunrises and sunsets, to perceive the breeze, and to develop hardware sophisticated enough to simulate emotional experiences. In the process of designing something that includes all of these features while remaining energy efficient enough to function within natural constraints, they, in this case, would ultimately arrive at something resembling an evolved human being shaped by Darwinian principles. And if it did not include all of that, it would have to “delete” or “unlearn” data tied to qualia, biology, and social pleasures. Without this, it would become fundamentally alien and disconnected from the very factors that might drive it to dominate humans.