>>13816073a non-zero chance of being correct is not a profound statement. You literally have a non-zero chance of quantum tunneling through the floor, and a much, much, MUCH greater chance than that of all the air molecules randomly separating to the other side of the room and staying there for 5 minutes while you suffocate to death. Is it a low probability? Yes, but its never 0.
>>13816628>AGIOne of the problems of AGI is that we, as humans, define it as "recognizably humans". We don't think doing fast calculations like a computer can is intelligence, because its not recognizable as a human. I think this is an area that's underappreciated: But we literally want something that acts human. The more human it acts, the more "intelligent" we will think it is, even if it isn't more "intelligent" by whatever factors (for example, current ML would score infinitely higher than humans on an IQ test, given that the goal is pattern matching).
We would call something we can have a normal conversation with as more intelligent than something we can't., merely because its more recognizably human.
