I think the Sequences got everything right and I agree with them completely.
Even Eliezer considers them first drafts (which is, after all, what they were: a two-year exercise in writing raw material for the forthcoming book), not things to be treated as received wisdom.
I think it was cool of Grognor to make a meta+ contrarian post like this one, and it’s a good reminder that our kind have trouble expressing assent. As for me, I see some of the sequence posts I see as lies-to-children, but that’s different from disagreeing.
As for me, I see some of the sequence posts I see as lies-to-children, but that’s different from disagreeing.
Could you elaborate on this? I interpret ‘lies-to-children’ as meaning that a model is too basic and is wrong in some places because there are applicable details which it does not take into account—would you not disagree with such things, if you don’t think that such models actually form a correct map of reality?
Well, I see it more like teaching someone about addition, and only covering the whole numbers, with an offhand mention that there are more numbers out there that can be added.
I meant in the sense of lies-to-children, or Wittgenstein’s Ladder. I cannot remember the primary posts that gave me that impression, and I know they were better than this concrete example; but that shows what I was talking about. At a sufficiently granular layer, technically incorrect; but inarguably useful.
Even Eliezer considers them first drafts (which is, after all, what they were: a two-year exercise in writing raw material for the forthcoming book), not things to be treated as received wisdom.
But, hey, maybe he’s wrong about that.
I think it was cool of Grognor to make a meta+ contrarian post like this one, and it’s a good reminder that our kind have trouble expressing assent. As for me, I see some of the sequence posts I see as lies-to-children, but that’s different from disagreeing.
But apparently we can hack it by expressing dissent of dissent.
How many levels of meta can we go?
Too many.
Could you elaborate on this? I interpret ‘lies-to-children’ as meaning that a model is too basic and is wrong in some places because there are applicable details which it does not take into account—would you not disagree with such things, if you don’t think that such models actually form a correct map of reality?
The entire Quantum Physics sequence is a protracted lie to people who can’t or don’t want to do the math.
This was explicit in the introduction to the sequence.
Well, I see it more like teaching someone about addition, and only covering the whole numbers, with an offhand mention that there are more numbers out there that can be added.
Drastic simplification, yes. Lie, no.
I meant in the sense of lies-to-children, or Wittgenstein’s Ladder. I cannot remember the primary posts that gave me that impression, and I know they were better than this concrete example; but that shows what I was talking about. At a sufficiently granular layer, technically incorrect; but inarguably useful.