User talk:Criptoriga

A universal human Turing machine is a person able to process symbols in a machine-like way, with minimal thinking and without interpreting them. The thought experiment of a human Turing machine starts from the question: what would it be for a human to pass as an AI, albeit a primitive or extremely limited one? The original Turing test's objective was to acknowledge the possibility of an AI simulating human intelligence. The experiment of a human Turing machine, on another hand, is supposed to show the possibility of a human being simulating credibly, consistently, a machine. The question that underlies this experiment is: could a human pass as a symbol processing software, in certain circumstances? The suggested answer: any human being able to repeat, copy or, generally, reproduce symbols in a way suggesting a mechanical approach to processing language in the absence of any interpreting goal and of an understanding effort should be considered a candidate for this kind of "machine". Any human who could pass as an imperfect Xerox machine from time to time will aproximate it. A revealing variant of the test could be to convince anyone reading a paper assembled by the candidate human Turing machine that it's the product of a primitive AI specialized in migrating large sequences of text from a group of sentences named, conventionally, "book" or "article", to a writing assignment of any kind in an academic context. A universal human Turing machine would process text like genetic code, introducing mutations randomly, altering meaning and the legibility of the text. This should be the proof of the lack of an understanding effort of any kind.