Ask HN: Experiences with 'Gödel, Escher, Bach' (GEB)?

Frankly it made me feel like I'm not smart enough much of the time. But trying not to let that bother me, I plowed ahead anyway. I found quite a few things actually quite mind expanding. His exposition about infinite recursion was worth the price of admission alone. A small segment of that is here: http://amberbaldet.com/uploads/little-harmonic-labrynth.html

Another book which I did find truly transformative include "The Art of Game Design" by Jessie Schell. Also Anathem, by Neal Stephenson, is also a very worthwhile mind-broadening scientific read. A New Kind of Science by Wolfram is also quite a bit of fun.

But by far my most prized techie book is Genetic Programming, by the legendary John Koza (if you're not a Lisp person like I wasn't when I read this, then you really need to read it!). Braid and all of the books I mentioned have something in common: They are all fat, heavy, meaty, thought-provoking tomes.

Great book. Like most people that start it, I didn't finish. It is a notoriously underfinished book. Give it a try anyways.

The first time I read it straight through. I was in the middle of college, and transitioning from a dual bio/chem major to a math/cs major, and the book really hit a sweet spot for me with the nature of the universe.

If you're a programmer, then you have a big leg up on the core concepts introduced in the first half, but its still a tough read.

The biggest piece of advice: keep going, and don't be intimidated as the examples grow. There are examples that are drawn from many fields of study, and the dialogues in between chapters give a more intuitive sense of the concepts being introduced in the following chapter. Re-read those if you start finding the main discussions too dry.

There is a _lot_ going on in that book, and you will not get it all on the first read. You will have a feeling that you are missing something and you are. Don't worry about that. Some is revealed in the later parts of the book, and you might find yourself looping back to earlier passages.

Take your time with it, but get through to the end, you will not regret it, and you will find yourself picking through parts of it for a long time to come.

Also, check out Metamagical Themas, which is the opposite, a series of discrete chapters on different topics, but just as fascinating. It covers lisp, rubics cubes, the prisoners dilemma, gender roles in language and a whole host of other good things.

I've read it two times through. The first time, I thought it was one of the most brilliant books I had ever read (I read it the first time in the early 80's). I became especially fascinated with Godel's Proof and the thoughts of E.O. Wilson on whether an ant colony was in reality a single organism. I was very, very excited about the possibilities of artificial intelligence and especially natural language processing.

The second time I read it (somewhere in the 90's), I enjoyed it tremendously but I was surprised by how excited I had been before. The second time through, I caught a lot of additional insights but somehow, it felt less revolutionary.

I plan to read it again in a few years. It's been over ten years since the last time I read it. I have no idea how I'll react to it.

I made an attempt to read Hofstadter's more recent I'm a Strange Loop about a year ago and didn't make it all the way through.

After being told (a few weeks ago, actually) that GEB was a bit more accessible, relative at least to ISL, I started reading it, and have thus far been blown away.

I highly recommend Gödel's Proof (Nagel and Newman) as a supplement. I got a lot out of Hofstadter's book, but honestly found much of the discussion of the incompleteness theorem more confusing than it needed to be. If you're mostly interested in the Gödel part, I actually recommend you read it instead of Hofstadter's book.

Gödel's Proof is very concise, and more clearly describes the core argument.

I read the last ~1/3 of it while recovering from food poisoning in a Buenos Aires hotel room. The simpsons was playing on TV, in spanish. Although the context was trippy, the content seemed pretty reasonable to me. It made a lot of sense from a theory of computation perspective; my response was mainly "oh, just like turing machines." Of course, I may have missed something.

I absolutely love this book - I'm right in the middle, he's trying to answer the question, "Can minds be mapped onto each other?".

It's very engaging and fascinating, and I highly recommend it for all sentient beings. Admittedly I haven't finished the book just yet, but I hold the belief that I may have been better off not going to college, but instead reading GEB two or three times.

One reason I occasionally stumble while reading it is because it's such a vast flow of information. It's a bit hard to process all of his thoughts at once, as it's basically a well-organized dump of Hofstadter's brain. Reminds me of Pirsig's Zen & the Art of Motorcycle Maintenance in that way.

As for how to approach reading the book, I recommend reading it outdoors in the summertime, which is a bit tricky in the northern hemisphere these days. And don't forget to have a LOT of coffee on-hand!

enjoy! :)

My major was Cognitive Science, so I always had to read excerpts of this book as part of course requirement, but never finished the whole thing.

Now I read it by jumping between pages and hoping the individual sentences/paragraphs could spark my own thought process (this is how I read most of the time nowadays).

For a much more concise overview on the similar subject matters, checkout this book by Andy Clark http://www.amazon.com/Mindware-Introduction-Philosophy-Cogni...

I hated it. I felt like the author was just trying too hard. There's just not enough actual content in a book of 800 pages.

As someone who does the majority of their reading in bed, GEB took a long time to get through, and I definitely didn't pick up nearly as much of it as I should have.

That's not to say it's necessarily dry or boring, it's just, I imagine your brain has to be on to really appreciate it.

I finished it, but found it pretentious, boring as hell and thin on content. In my opinion (which is not shared widely ;-)) it would have made a good 15 page pamphlet. The rest is turgid verbiage(imo). I can understand someone for whom recursion(say) is some kind of mind blowing concept getting some value out of the book, but I've never understood why programmers find it fascinating. It has a few ultra basic concepts, stretched across multiple pages with lots of padding, then endlessly repeated in minor variations with lots of faux insight.

I've read it twice. That says it all, I guess.

I thought it was awesome when I first read it in high school; the second time, after graduating university, I skipped all of the Achilles & the Tortoise sections and became quite frustrated at how many page turns it took to contain any given idea. Like Devilboy said, there's not reason it should have been 800pp.

One of the few books I've put down after less than 1/3rd. Really bad and over-hyped, IMHO.

I first read GEB 2 years ago and was similarly mind-blown. I found that it didn't answer any questions, but it did ask a lot. I would consider essential reading for anyone studying AI, and indeed for anyone with a curious mind.

I'm also in the middle of I Am a Strange Loop, and it feels very familiar. Hofstadter's wit and skill with words comes through strongly, and the technical depth of thought is equally as compelling.

With regards to approaching the book, I found it enjoyable and helpful to stop every few pages and think about some of the concepts that are posed. Don't consider it a book on AI, or computing, or Bach, or Escher, or Godel, or quirky narratives. Instead, just go with it and follow along with the story. It takes a while, but once you've finished, you'll want to read it again.

I love it. But:

The book is NOT a good introduction to the incompleteness theorem. It is not a book about M.C. Escher. It could, barely, be said to be about papa Bach.

If you read it for math or science, you will be disappointed. It's a horrendous textbook.

Read it paying attention to his words, how and why he choses them, and GEB is a pretty marvelous thing. If the English language excites you, then read it, and read Le Ton Beau de Marot. GEB is a demonstration of ideas, not a discussion of them.

Hofstadter is not a computer scientist. He's a philosopher, a linguist, and a gedankenspiel-smith. Read him to laugh and play. Expect pretension because big words are fun.

Or skip it. I don't think it will change your life.

Yes, probably the single most important book I've read, and by far the best tutorial in symbolic logic that I know. Best approach: disappear for a month, take a holiday from the internet.

I'm in the middle of it and reading it slowly. It hasn't quite lived up to the hype, but I have deeply enjoyed it. I think it's wonderfully quirky and while yes, it's probably too long, I think that's ok, because it's not a blog post and I am happy to make room for something massive and rambly on my bookshelf.

Good fun but too long. I loved it when I first read it, and in college I ran a reading course where we read it over one semester. But when I learned more and later came back to it, I found it disappointing content-wise. I'm still glad I read it, though; it was pretty formative for me.

GEB is a great experience. It's not a deeply technical book, though it is both deep and technical. It's more specifically an adventure of thought.

I hear there are hidden messages in the book itself. I'm not motivated to track them down and solve them, but this is certainly an indication of the kind of tone Hofstadter was after: playful.

It got me hooked on programming, classical music and Escher at a young age. The best way to approach it is to realise the author's intention. He wants to explore the question: what is a self? If you keep that in mind, then his overall approach makes more sense.

I am 200 pages in. For now, it has just radically changed my view of the concept of formal systems.

You should read GEB like a playful, mathematical Alice in Wonderland full of puzzles, philosophy and sometimes self-indulgent wit. Once you go down Hofstadter's rabbit hole you'll get a mind-expanding romp through the meta-logical underpinnings of modern mathematics, information theory and computer science. At the end, whether you find it deep or simultaneously twee and pretentious is a matter of taste, but you'll definitely find yourself a lot more capable of thinking of three impossible things before breakfast.

Based on what I'd heard, I expected a lot more from it. At times, Hofstadter somehow manages to turn what should be insanely interesting subject matter into something quite dry.