Raw LLM Responses

Rule based AI the way you explain it isn't AI; it is software with some if statement that even a primitive programming language must be able to do to have any use at all beyond of that of an addition machine. Rule based AI is actually a much more advanced version of what you have mentioned and we call it fuzzy logic in which it combines various different parameters to produce a result. For example some cars of the 90's would have fuzzy logic ECU's to appropriately adjust various functions such as injector and timings and A/F mixture as well as changing gears in automatic cars. For example it would gather information via various sensors such as vehicle speed, wheel speed differential, engine RPM, tractive load and throttle position (as well as dozens upon dozens more parameters) to decide whether to shift up, shift down or keep it in the same gear. While it is in essence a series of if statements to put it into perspective it is a much more complex decision tree which emulates a reasoning process (without having reasoning capabilities in itself). Self aware AI will not come after super-intelligence is achieved but most probably shortly after AGI. Chat GPT isn't yet a reasoning AI as its reasoning is based on emulating intuitive predictive algorithms based on huge data rather than real problem solving abilities. We are now on the cusp of reasoning AI which will lead to AGI eventually; to have a reasoning AI it must first be able to comprehend mathematics (not just spew answers but really understand it) and now the next iteration of AI models, at least from OpenAI, will be able to do just that. Now there's no point of watching the video beyond that point because beyond Super-Intelligent AI we cannot fathom the direction that AI will take. It is mere human speculation based on our imagination and instincts alone.