How Gödel's Proof Works (2020)

Also recommend this book. Very concise too, it's only around 100 pages.

Wish I had the full text on me, but I recall there's a version of it with an introduction by Douglas Hofstadter that disagrees with the interpretation of the proof by Nagel and Newman, and it's something key to their culminating assessment of what Gödel's proof truly means, which is a remarkable disagreement to put into a forward to a text. Unfortunately the Kindle version I got from Amazon doesn't contain his forward and I'm struggling to find that version.

Edit: found the quote, and frankly I agree wholeheartedly with Hofstadter who seems to have a much more sophisticated understanding of the capabilities of computers, perhaps owing to his encountering them in a later era.

"My book, despite owing a large debt to Nagel and Newman, does not agree with all of their philosophical conclusions, and here I would like to point out one key difference. In their “Concluding Reflections,” Nagel and Newman argue that from Godel’s discoveries it follows that computers—“calculating machines,” as they call them—are in principle incapable of reasoning as flexibly as we humans reason, a result that supposedly ensues from the fact that computers follow “a fixed set of directives” (i.e., a program).

To Nagel and Newman, this notion corresponds to a fixed set of axioms and rules of inference—and the computer’s behavior, as it executes its program, amounts to that of a machine systematically churning out proofs of theorems in a formal system. This mapping of computer onto formal system takes the term “calculating machine” very literally—that is, a machine built to deal.with numbers and arithmetical facts alone. The idea that such machines by their very nature should churn out sets of true statements about mathematics is seductive and certainly has a grain of truth to it, but it is far from the full vision of the power and versatility of computers.

Although computers, as their name implies, are built of rigidly arithmetic-respecting hardware, nothing in their design links them inseparably to mathematical truth. It is no harder to get a computer to print out scads of false calculations (“2 + 2 = 5; 0/0 = 43,” etc.) than to print out theorems in a formal system. A subtler challenge would be to devise “a fixed set of directives” by which a computer might explore the world of mathematical ideas (not just strings of mathematical symbols), guided by visual imagery, the associative patterns linking concepts, and the intuitive processes of guesswork, analogy, and esthetic choice that every mathematician uses.

When Nagel and Newman were composing Godel’s Proof, the goal of getting computers to think like people—in other words, artificial intelligence—was very new and its potential was unclear. The main thrust in those early days used computers as mechanical instantiations of axiomatic systems, and as such, they did nothing but churn out proofs of theorems. Now admittedly, if this approach represented the full scope of how computers might ever in principle be used to model cognition, then, indeed, Nagel and Newman would be wholly justified in arguing, based on Godel’s discoveries, that computers, no matter how rapid their calculations or how capacious their memories, are necessarily less flexible and insightful than the human mind.

But theorem-proving is among the least subtle of ways of trying to get computers to think[...]"

He goes on like this for a bit more, and fleshes out a deeper argument, but this is already long as quoted passages on hn go. But I think Hofstadter is exactly right and shows a much more sophisticated understanding of computers than Nagel and Newman in their celebrated introduction to Gödel. I would go so far as to say their philosophical conclusion is almost exactly wrong, and is as baffling as if Darwin's Origin of Species included a section of "conclusions" denying the possibility of ever developing effective vaccines in the future. Wrong to the point of being contrary to the spirit of the subject that was so exceptionally articulated up to that point. And it's in my opinion terribly damaging for a conclusion so backwards to be embedded in a text that's celebrated as the best explanation of the proof.

GEB is a great book, and I've probably read it at least 3.33333333... times over the years. As a late teen it blew my mind. But I'm not sure I'd recommend it as a route into Gödel's proofs [1]. The book covers a lot of other ground too, and is notoriously digressive and quirky (looking at you, dialogues).

Instead I'd recommend Gödel's Proof by Nagel and Newman for a conceptual intro.

[1] I'm not a mathematician, so my understanding is necessarily informal.

“Godel’s Proof”[1] is also a great and shorter read if the scale of GEB is intimidating(I know it was for me at first).

https://nyupress.org/9780814758014/godels-proof/

It’s actually an interesting fact that every person who was programming in the 80s owns a copy of GEB, which they flipped through a bit and then put on the bookshelf and never actually read.

while it is a groovy into to recursion and other cool ideas, GEB annoys me in that I feel like the three figures in the title are ill matched. Godel proves a super important result in math, sure... Escher was a skilled draughtsman who had a feel for tesselation. An OK artist IMO but no special insights. Bach on the other hand was an expressive genius who in the volume, power and beauty of his productions just seemed to drop out the sky like a meteor. Escher does not belong in the same breath frankly. if Bach made a crab canon or did other marginally math-y things that is just not the point - the work lives or dies in entirely different terms...

I hope you don't mind, but just in case anyone is curious like I am, here i think is the video to the talk https://www.youtube.com/watch?v=KdZq5JvhPVQ

Very Nice; Thank You!

Is there a way to get a pdf of the slides?

It’s clearly true in The Natural Numbers. It’s not provable because in some model it’s false. Being clearly true in one model does not make it provable.

Thank You.

A classic Mir publication from the Soviet-era.

That won't work. Gödel number encodes a paradox, but a halting execution of a TM is not a paradox, so can't be written as a Gödel number.

Does your uncountable set of propositions include some which cannot be written as a finite string of symbols?

If "yes" - how might such as proposition be proven true with a finite string of symbols?

(If your proof symbols are from an infinite character set, that has its own issues.)

Even more videos to binge before the holidays end! Thanks for sharing!

Nagel & Newman is much better if you find GEB a slog (which was also my experience). However, Gödel’s theorem is fundamentally due to self reference. But so is the fact that the power set of countable infinity is uncountable, in a similar way.

I think you certainly grasp the gist of things if you understand that the point is self-reference. What makes the proof shocking and beautiful (at least, IMO), is how sparse a toolbox Gödel is working with (and, perhaps, how clever the construction is). Contradictory self-reference in, say, plain English or naïve set theory, is not particularly surprising (at least to our modern eyes), because those are rich languages where you can express an extremely wide range of statements, but Gödel manages to smuggle it in using just whole-number arithmetic and first-order logical statements, which provides a recipe for doing so in essentially any formalized system of interest to "normal" mathematicians.

Some comments elsewhere in this topic mention that you can use the halting problem to arrive at the same essential conclusion in a more understandable way (and certainly one less fraught with small technicalities to work through), but (again, IMO) there's a certain mischievous magic to the way the Gödel sentence gets constructed that makes it very fun to work through for the first time.

I thought it was sorting an infinite set of infinite strings as the first step in the "algorithm" that seemed sketchy. [0]

It's definitely not a constructive proof, even though it pretends to be; none of the mathematical objects can be constructed, nor can any of the algorithmic steps be executed.

That said, it's far less well-known that the workaround (if you find it to be true) is trivially easy (from Alfred Tarski), making it kind of a useless theorem in practice.

[0] You might think, well, I'll just write out the strings in order by using a generator! No sorting needed... But you have to write the strings down to perform the algorithm, and it takes infinite time to write down the first string, so you'll never even get to the others which is when you do the diagonalization trick. Like I said: it's not constructive, none of it can actually be done.

There is no infinite regression involved in any of Godel's theorems.