3 ms·

I'm not sure why anyone would consider it out-dated. None of the concepts are really things that are 'datable'. That said, you'll probably find that you need t

by serf 13y ago

I'm not sure why anyone would consider it out-dated. None of the concepts are really things that are 'datable'. That said, you'll probably find that you need to read it more than once to glean everything from it.

slashnull 13y ago
I have to say I was surprised by that statement, but, thinking of it... Category theory and Type theory are replacing Set theory as the fundamental basis of Math, or at least it's fashionable to try to do so, and the book revolves around the Whitehead and Russel set theoretic work and Godel's deconstruction thereof. Which isn't to say that Category theory or whatever are immune to Godel's theorems, I have no idea how those would translate, and my gut tells me that they would have roughly the same outcome. But yeah, anyways, wonderful exposition of the kind of extremely bare-bones framework fundamental mathematicians operate in, magnificient demonstration of what recursive and self-referential structures imply, and overall a great read.
spamizbad 13y ago
One thing that stood out during my reading was Hofstadter's speculation on the reproduction of viruses. We have a much better understanding of that now and it doesn't match his speculation. Thats the only thing that comes to mind and it's only a minor quibble.
rspeer 13y ago
When he writes about AI, he's writing about "good old-fashioned AI" -- the kind that was supposed to be advanced by clever representations, recursive data structures, and layers of abstraction. Contrast "good old-fashioned AI" with "machine learning", where the representations are as minimalistic and low-level as possible, because the only important thing is that when you update them a few billion times the right results emerge statistically. I don't see it as a problem that this part of GEB is "dated", though, because I find GOFAI to be an interesting topic. It involves trying to think about how you think and formalize it, which rarely happens in modern AI. Imagine a world where brute-force alpha-beta search was just not good enough to beat humans at chess; a world where advances in chess-playing computers would require chessmasters to encode their expertise in interesting data structures. This is the world that might have been, and it's the world GEB describes when it talks about AI. It's dated, but it's interesting.

by serf 13y ago

slashnull 13y ago

I have to say I was surprised by that statement, but, thinking of it... Category theory and Type theory are replacing Set theory as the fundamental basis of Math, or at least it's fashionable to try to do so, and the book revolves around the Whitehead and Russel set theoretic work and Godel's deconstruction thereof. Which isn't to say that Category theory or whatever are immune to Godel's theorems, I have no idea how those would translate, and my gut tells me that they would have roughly the same outcome. But yeah, anyways, wonderful exposition of the kind of extremely bare-bones framework fundamental mathematicians operate in, magnificient demonstration of what recursive and self-referential structures imply, and overall a great read.

spamizbad 13y ago

One thing that stood out during my reading was Hofstadter's speculation on the reproduction of viruses. We have a much better understanding of that now and it doesn't match his speculation. Thats the only thing that comes to mind and it's only a minor quibble.

rspeer 13y ago

When he writes about AI, he's writing about "good old-fashioned AI" -- the kind that was supposed to be advanced by clever representations, recursive data structures, and layers of abstraction. Contrast "good old-fashioned AI" with "machine learning", where the representations are as minimalistic and low-level as possible, because the only important thing is that when you update them a few billion times the right results emerge statistically. I don't see it as a problem that this part of GEB is "dated", though, because I find GOFAI to be an interesting topic. It involves trying to think about how you think and formalize it, which rarely happens in modern AI. Imagine a world where brute-force alpha-beta search was just not good enough to beat humans at chess; a world where advances in chess-playing computers would require chessmasters to encode their expertise in interesting data structures. This is the world that might have been, and it's the world GEB describes when it talks about AI. It's dated, but it's interesting.