Raw LLM Responses

Neural network principles have existed since the analogic era, it's nothing new. The problem has always been, and still is processing power; a lot of binary checks necessitates exponential amounts of energy, which we still cannot afford for AI neural networks to this day. Even recent AIs are at a net deficit, aren't even profitable (ask Sam Altman) and can only manage one sense, since they're just parsing binary data. Meanwhile the supercomputer that is our brain is extremely compact, can sense in multiple dimensions in addition to simple visual parsing (sound, touch, smell, depth, taste) and is also able to layer them one over another to make complex decision making in a snap, and is also able to store data using different priorities (Short, Long, Subconscious, etc) doing so at an efficiency that eclipses what modern AI can ever do (Most LLMs run out of memory in a matter of 200 paragraphs or so, for a room sized server, it's ridiculously small). All of that at the price of only a few calories daily for both itself and its motor abilities. Truly AI is very far from reaching effective workflow yet alone consciousness by thousand of years sadly, and it's less a knowledge issue of how to make complex AI, and more about how to make it much more energy efficient and more compact to manageable levels. The only reason it boomed lately is because some people found a way to 'cheat' AI into giving an acceptable result using noise-based parsing, which saves energy at the hefty cost of accuracy, it's more of a magic trick than anything truly useful, because noise based solutions are always a 'good enough' approximation, which when coupled with the instruction to always give an answer to everything with non-null returns makes them very prone to hallucination, which can be summed up by 'inaccurate + always right'. The later is probably going one of the hardest problem to tackle, and has been in electronics for a while now; how do you compute the absence of something, AI knows what it knows, but it cannot quantify what doesn't exist, it's complicated.