The stone age analogy doesn’t quite fly. There’s a difference between the state where I want X and someone else is offering X, and the state where nobody is offering X.
But you’re right, of course, that even in the first case I have to know I want X.
That said, I don’t have to know the name of the field.
For example, if I’m genuinely interested in what actually happened in ancient Rome (which is of course only one possible meaning of the phrase “ancient Rome history”), I will sooner or later discover that there are disagreements among experts about some aspects of it, as well as questions about it that are simply unanswered.
A genuinely curious person, who is actually motivated to know what actually happened in ancient Rome, will in consequence sooner or later have thoughts like “how do I decide which experts to believe?” or “how do I decide which of these competing theories is true?” or “how do I come up with answers to questions that haven’t been answered yet?”
There’s no particular guarantee that it will occur to them that the thing called “probability theory” or “cognitive science” is related to that (though “decision theory” seems like a reasonable thing to investigate), but asking those questions cogently enough of the Internet, often enough, will sooner or later get them there, assuming they don’t settle for something along the way that doesn’t actually satisfy their curiosity but gives them some other thing instead.
I think you overestimate the ease it is to “jump to the meta-level” (ie, you want to learn about something, so you jump to learning how to learn) to people who were not pointed to do it—by reading Gödel, Escher, Bach, some of LW or anything like that. Someone genuinely curious about “what actually happened in ancient Rome” will read lots of books about it, will go to visit the ruins, go to museums, … but won’t spontaneously start asking about “decision theory” or about “what is the general process to resolve dispute between scholars ?” if not given strong hints that they should do it.
I don’t think I’m overestimating the ease; it’s difficult. And, yes, it generally requires some sort of outside input.
In practice, though, the reason people don’t do this isn’t because they couldn’t do it with, say, a year or four of consistent asking, receiving answers, evaluating those answers for whether they actually resolve their genuine curiosity, discarding those answers which don’t, and repeating the process.
The reason they don’t do it is because they settle for one of the answers they get early on, and stop asking.
The stone age analogy doesn’t quite fly. There’s a difference between the state where I want X and someone else is offering X, and the state where nobody is offering X.
But you’re right, of course, that even in the first case I have to know I want X.
That said, I don’t have to know the name of the field.
For example, if I’m genuinely interested in what actually happened in ancient Rome (which is of course only one possible meaning of the phrase “ancient Rome history”), I will sooner or later discover that there are disagreements among experts about some aspects of it, as well as questions about it that are simply unanswered.
A genuinely curious person, who is actually motivated to know what actually happened in ancient Rome, will in consequence sooner or later have thoughts like “how do I decide which experts to believe?” or “how do I decide which of these competing theories is true?” or “how do I come up with answers to questions that haven’t been answered yet?”
There’s no particular guarantee that it will occur to them that the thing called “probability theory” or “cognitive science” is related to that (though “decision theory” seems like a reasonable thing to investigate), but asking those questions cogently enough of the Internet, often enough, will sooner or later get them there, assuming they don’t settle for something along the way that doesn’t actually satisfy their curiosity but gives them some other thing instead.
I think you overestimate the ease it is to “jump to the meta-level” (ie, you want to learn about something, so you jump to learning how to learn) to people who were not pointed to do it—by reading Gödel, Escher, Bach, some of LW or anything like that. Someone genuinely curious about “what actually happened in ancient Rome” will read lots of books about it, will go to visit the ruins, go to museums, … but won’t spontaneously start asking about “decision theory” or about “what is the general process to resolve dispute between scholars ?” if not given strong hints that they should do it.
I don’t think I’m overestimating the ease; it’s difficult. And, yes, it generally requires some sort of outside input.
In practice, though, the reason people don’t do this isn’t because they couldn’t do it with, say, a year or four of consistent asking, receiving answers, evaluating those answers for whether they actually resolve their genuine curiosity, discarding those answers which don’t, and repeating the process.
The reason they don’t do it is because they settle for one of the answers they get early on, and stop asking.