Matt: 0:14

Hello everyone And welcome to the switch statement podcast It's a podcast for investigations into miscellaneous tech topics

Jon: 0:21

This is our ninth episode on Gödel, Escher, Bach by Douglas Hofstetter.

Matt: 0:31

Hey, John, how are you doing?

Jon: 0:33

Hello, Matt. I am doing good. how are you doing?

Matt: 0:37

I am doing all right. I'm ready to learn about the location of all meaning.

Jon: 0:44

Yeah, where's the meaning?

Matt: 0:45

My search for meaning ends today. I'm going to find it. If he tells us the location of meaning.

Jon: 0:51

Yeah, yeah, this chapter is it, people. It's like, this is where meaning is. This is one of the toughest reads we've done, but.

Matt: 0:59

you're saying, or this

Jon: 1:01

This book, this book. in general, uh, but it's more so because the, the material is just so complicated.

Matt: 1:08

Although, I will say when I was scrolling through my PDF that I'm reading off of. Like where we're a quarter of the way through. Which surprised me a little bit, because like, yes, the book is very dense, but there's a lot of, I wouldn't say fluff, but like those dialogues are, are adding a lot of. Heft to the book.

Jon: 1:29

that's crazy. They were a quarter of the way through. I'm kind of proud of that. Cause I mean, I have a physical book. and it is a straight up tome. Like it's a large object.

Matt: 1:39

Yeah. Well it's because he's, he's laying down a lot of these, Interesting examples, like where does meaning live? Inside the message or inside the brain of the person?

Jon: 1:55

Yes. Yeah. So he kicks this. chapter off as always with one of his weird, uh, what does he call them? Dialogues.

Matt: 2:04

Yeah. Which I didn't read. I don't know if you read it and maybe we just keep on going back and

Jon: 2:09

so I did read it this time because the last, you know, the last chapter had the dialogue and then the chapter like refer to the dialogue a trillion times. This was different. I don't, I mean, I think the chapter alluded to the dialogue a few times, but like very vaguely, but this dialogue, I will give a quick. Description of the dialogue because it did relate to the chapter in an interesting way. So the tortoise and, uh, is it agent or no Achilles? I'm set Ajax, uh, they're hanging out and they find this jukebox, but the way the jukebox works is it only actually has one record, but that record gets put onto different record players and depending on the record player, it plays something different. So the examples they give is they put it on one record player and it plays our song Bach, B A C H, even though H isn't a note, but they did mention that in German, in Germany, H was a note for some reason. Um, but anyway, they, they put that same record on a different player and it plays C A G E, which is another composer named John Cage that I think we've actually talked about before. I think we actually mentioned 4, 433. But in any case, that's

Matt: 3:27

But is that just silence? There's just nothing. Ah,

Jon: 3:31

well, no, I mean, he just, he mentions that at one point, but that's not what the record plays. The record has four notes, but the point is like each record player almost represents a way to map those notes into music, you know, so the record has some data on it that's going through some, you know, you can think of it as like a formula. Which is the record player. And then the output is the music, which is different depending on the record player.

Matt: 4:00

Right.

Jon: 4:01

So I think it, I think it's meant to just kind of introduce this concept of having like underlying data, but then having like a decoder that actually results in information.

Matt: 4:12

I'm really glad you brought this up and maybe I'm jumping ahead too much. He makes a claim that. Uh, jukebox is not a decoding mechanism. Which. I found very, like, I'm sure there's a subtlety that I'm missing there, but, um, But I guess the way I interpret what an, what a decoder is is there's something in one form. And then it gets like, Translated into another form that life is more useful.

Jon: 4:43

Yeah. Yeah, exactly. He sort of kicks the actual chapter off talking about information bearers. And information revealers, which this is like, you know, a very simple way of talking about this. I feel like, you know, he does this classic thing where he starts really simple, almost with a goofy story. And then he gets into actual, you know, information, but he still starts simple. And Yeah. he talks about the record itself, you know, with the physical grooves being the information bearer. But then the, the player is, uh, the information revealer. He also brings it back to the, PQ system though, he mentions again, where he talks about the theorem being the bearer, but the interpretation is the revealer, which I thought was interesting. So I guess the way you interpret the theorem. And you know, if we recall the, the PQ system was like. You know, two dashes P, two more dashes Q, four dashes. So it's basically simple arithmetic. But it's like, that interpretation that it is simple arithmetic is the information revealer in this case, I guess. At least that's how I read this part.

Matt: 5:59

Yeah, I, I just kind of glossed that over. I read that section and I wasn't a hundred percent sure. What he meant by that. And even upon further reflection, I think I, so don't really. I'm just not sure I buy it.

Jon: 6:12

I, I had the same exact feeling. But then, Like he mentions later in this chapter that there is this notion of decoding, where you have like raw information, and you sort of decode it, and you end up with, you know, new information, but based, but depending on the. decoder, your interpretation of the result might be different. well, I guess put more simply, it's like, depending on the recode, the decoder, the result is different. So you can get vastly different results based on the decoder you're using. Yeah,

Matt: 6:50

decoder argument. And this is why he's saying maybe that a jukebox is not a decoding mechanism. Is, this is a case where there is actually like an underlying answer. And all decoders would. Resolve the same answer. cause this gets back to the original question, which is like, does meaning live in the object or does it live in the interpretation? And like, I think his point is like, In cases where the information lives in the object itself, like. That's the context when you can have a decoder and a decoder will always resolve to the same. Answer. Which proves that like there there's, there's a truth. And then underlying truth to the. I would like, or there's an underlying information in the object. Um, does that, does that align with how you're interpreting it?

Jon: 7:49

Uh, well, so I'm not sure because I definitely interpreted it as though, you know, there is a Right. there is a correct decoder, but there is also, You know, part of deciphering information is like knowing the correct decoder because there's also like incorrect decoders, which you're obviously not getting the intended meaning out of the original message if you're using the incorrect decoder. But that that was sort of a concept that I sort of understood from this chapter is like, you know, you have this raw data. Uh, or this raw message. Which it may or may not be obvious what the correct decoder is. And there's a lot of discussion in this chapter about messages where like the decoder is very obvious given the message. Um, but in any case, it's like in order to truly decipher a message, you need the raw data, you need to know the correct decoder, and then you need to apply that correct decoder to the raw data. Um, and actually just to make sure we're not getting too, like, Uh, vague about all this. Um, he mentions another example, which is like a molecule of DNA where, and DNA has, you know, these terms where you start with a genotype and you convert into basically a functional physical organism called a phenotype, and this process is called epigenesis. And he talks about how. You know, this is an example of, like, you got the raw data, which is the DNA, which is essentially data, like it's, it's, it could be represented as binary on like a, you know, flash card or something. And then you apply some process to it, which is essentially physics, and you're left with an actual living functional organism.

Matt: 9:47

Right. And he calls this an exotic isomorphism. Uh, because. It's it's completely. Unclear how, a unit of DNA is a mapping onto a unit of the organism. Right. It's like, it's very, that does this very complex transformation. let me know if this was not your interpretation, but my understanding is That exoticism kind of lives in the complexity of the context. You know, it's like the reason why there's like a very unclear mapping between the. Uh, DNA. And the result is because like, there's this rich context that the DNA is tapping into.

Jon: 10:32

Yeah, Yeah. I, I definitely got that as well. I also, another thing I wrote down here was attention, uh, which, you know, in large language models, attention is this concept of like, you know, how strongly does one word in potentially a large body of text relate to another word? Um, you know, like let's say you have a sentence about, uh, the president or something. And it's a super long sentence, or maybe it's a whole paragraph, but the paragraph is introduced by saying, you know, something like this person is a liar. Like, that's an example where the word liar would have, like, a very strong influence on the meaning of the entire paragraph, because it's, like, a very charged, charged word, has a ton of meaning, and if you're talking about a president being a liar, then that's, like, a very impactful concept. But these two words could be, like Really distant from each other in the, in the data. And he talked about how, like, you know, in order to delete someone's nose, you can't just like delete a single value from the DNA, right? Like you'd have to manipulate tons of values across the entire, you know, DNA data, which I felt, I mean, at least to me, it's sort of related to this like attention concept where it's like, you have the concept of the nose, but it actually. Maps across the entire DNA data, but each, you know, there's tons of, there's tons of pieces of data across the DNA that like subtly map to how the nose ends up. Yeah,

Matt: 12:09

just point to The nose gene. Makes this an exotic. Isomorphism.

Jon: 12:14

exactly. It, and then, yeah, he mentions prosaic being the opposite of exotic, which goes back to our earlier example of like a record and music, you know, with a record, it's a physical object where you can, you know, You know, if you have a microscope, you can literally see grooves on the record and those grooves map directly to sound waves.

Matt: 12:35

Exactly and Nikes are where you can do this physical to temporal mapping between like the song and, and. The groove and like, so, so. Yes. If 17 hundreds Bach. And you've got a record. Like it's difficult to construct a system that does that, but once you have a system that maps like that mapping is direct pretty much.

Jon: 12:59

Right. Yeah. he mentions that the decoding of DNA is actually more complicated than the, than the data in the DNA itself. And I think you, you just sort of mentioned this, but it's like, in order to have a living, breathing human being. It's like, sure, you have this massive quantity of DNA, but then you like basically apply physics to that, which physics is like its own set of, you know, completely comprehensive and complicated laws of how the universe behaves. And so you have this, you know, you need to pass the data through this like utterly complicated decoding system in order to yield, you know, the actual result.

Matt: 13:43

I was also thinking. About the structure of the cell. Itself. Like. I think there's a bunch of those, those parts of the cell, which actually are not controlled or like, this is what I was, this is what I was also imagining. the process by which like the, yeah, the encasing of the cell. Grows. And what have you is not necessarily controlled by DNA is my understanding. And like this, this is also, I mean, Maybe this is just another facet of what you're talking about. The light. It just got me to think about how that, like there's a bunch of stuff about human reproduction that like, Isn't DNA, like mitochondrial DNA is like different than

Jon: 14:30

Oh, totally. Totally.

Matt: 14:31

you know? And, There's actually a lot there. There's a stuff you should know, episode, which was called like Kenya grandfather's diet. Affect you. And like, It has to do with the fact that like, Yeah. There's so many more contexts than just your DNA to consider. When you have to consider how a person is going to wind up turning out. Um, so yeah, this stuff's fascinating.

Jon: 14:56

uh, totally fascinating. Uh, he goes on to mention this concept of like, you know, and, and this, again, this is a lot about like information decoding, but he brings up like passing information to aliens and how, let's say you took a record and you wrote, or you put Bach's F minor, I think it was a violin sonata on it, would aliens be able to decipher that? And what that means is like, you know, what aliens look at the physical record object, they'd notice there's grooves on it. They might think, oh, we need some kind of device in order to play this thing. And then they might even hear like the Bach music once they find that device or construct that device. But even after hearing the Bach music, are they getting the message that we intended?

Matt: 15:45

D can they even like perceive sound? That's the other

Jon: 15:50

right.

Matt: 15:51

They may have evolved in a context where. Detecting differences in air pressure. Like it wasn't necessary.

Jon: 15:58

100%. Yeah.

Matt: 16:00

what's what's interesting is it's easy to be like, oh, they build an object that like, You know, Creates differences in air pressure, but it's like, there's no reason they would, like he's saying like, or he goes on to say this in the chapter. If the alien can't perceive sound. Through the air, like they're not going to build a device that. You know, cause it just w like it w it would never be useful. So they may find another mapping onto something that like they can perceive.

Jon: 16:30

Right. Exactly. Yeah. I recently listened to the John Williams Smartless episode

Matt: 16:37

Hm, what crazy they got John Williams

Jon: 16:40

They got John Williams who is an absolute legend obviously. But anyway, John Williams, he discussed this movie Close Encounters of the Third Kind. Which is an amazing movie, but the whole premise of that film is communicating to aliens through music. And there was, there was two things that John Williams mentioned that I just want to repeat. Uh, he mentioned that in the script of Close Encounters, he was tasked with coming up with a five note melody. You know, which ended up being, Bom, bom, bom, bom, bom. Like that famous Close Encounters melody. But he, he talked in the podcast about how five notes isn't typically a melody in human culture. Like five notes is typically something like a doorbell. Like it's like a chime. And then he, he mentions how seven notes is, is a melody, you know, like somewhere over the rainbow or something, which I just thought was super interesting. It's like such a cultural thing. But I started thinking in my head of like various melodies I know about. And it's, it's kind of true. It's like, you sort of need a certain amount of information before it starts to feel like a melody. Uh, and then he mentioned this other thing that I totally love. He mentioned how the way he works on music is he sort of like removes and prods things until it. feels like it was always there. Like to him, it feels like he's sort of subtly manipulating something until it just feels like this universal melody, almost like he's discovering like a math formula or something, um, which I just thought was cool. And, but it also implies this sort of universality of music, which is something that's discussed in this chapter, you know, like box music, it's very mathematical. It's, it's often. You know, it can often be described as like a series of intervals that are used, you know, for a reason. Some intervals cause tension, some intervals resolve tension. And so you can imagine kind of combining these concepts, almost to create like a mathematical formula. So in that way, maybe music is this universal thing.

Matt: 18:58

Yeah. I mean, something that always struck me about music. And these isomorphism between sound and. Music, I guess is the fact that. when you double the frequency of something. It sounds the same again Like, why does that happen? Like, It just, you like, you hear it and you're like, oh, well, that's just, you know, that's just how it is. But like, there's nothing to say that that's how it should be, you know? Uh, The fact that like, when you double the frequency. It Rhymes, I guess, in our ears in a way. Um, uh, you know, Hard to put words to how to notes that are like the same note. Sound alike, but,

Jon: 19:42

Yeah.

Matt: 19:43

that is a very interesting, interesting point. And actually, I mean, W D you know, keeping on that topic of John Williams, like I'm amazed that he is able to maintain. The distance from. Uh, thing that he's working on, because I feel like when I'm. Working on something. I just lose all ability to tell. it just becomes what it is. And I lose the ability to did make an like an objective assessment of it.

Jon: 20:09

Yeah. Yeah. John Williams is a very special person. And I don't know, I, I sort of imagined him having abilities that most of us mortals can't even conceive of.

Matt: 20:27

there was a very small point and He uses the word gross. The it's gross physical structure. Is very attention drawing. And this is that this directly ties to what we were talking about. Um, how, if an alien. Civilization detected this. Floating record. They would be able to be like, okay, it's pretty clear that this didn't just like randomly occur in the universe. Um, but so he describes. You know, it's gross physical structure and. I went on this little rabbit hole of like, trying to understand. The etymology of the word gross. I obviously have a couple of different like definitions for the word gross. This didn't really map onto any other ones that I knew. Um,

Jon: 21:11

means great in German. Yeah.

Matt: 21:16

means great in English too. Like that's kind of one connotation, like gross. Uh, profit or what have you like the big number, I realized, like, I actually didn't know the like main definition of gross. Like the fact that gross means like disgusting is, is. Is not in the top light, several definitions of the word. Uh, gross and like, it's this like informal meaning for the word gross, but, um, And listeners are probably aware of this, but. The way he's using gross here is like just immediately visible, like on its face. And that's like kind of the gross anatomy. Uh, connotation of the word gross, but,

Jon: 22:00

Right.

Matt: 22:02

I was very curious to know when we started to use gross in the light disgusting sense of the word. And I went on this like very long rabbit hole. Um, and the it's so funny because the first usage of the word gross, I could find the, like in that connotation. Is a book called, Limericks: Too Gross by Isaac Asimov and some other guy. So it's this, it's like 288 Memorex, like dirty limericks by Isaac Asimov and this other. Yes. Yeah. Um, no. I, uh, I know this sounded like it was going into the, into the weeds, but yeah. So I was just like, I didn't read any of them, but, um, and this was in 1978, I think that he came out with these, um, And so anyway, that was just a little random tangent, but that was one of the earlier like printed, uh, references to gross in the light. Disgusting. Meaning I defined.

Jon: 23:01

Wow. Isaac Asimov, man. He's like Shakespeare. He's just defining. Or changing the definitions of words.

Matt: 23:08

Riding dirty, dirty limericks in his spare time.

Jon: 23:11

Yeah. Got to make it dirty.

Matt: 23:14

All right. Well, so this might be a good, uh, stopping point. So why don't we, uh, yeah, it's a little cliffhanger. We'll we'll uh, leave, leave a little more. Uh, and you'll have to tune in next time for the. Dramatic conclusion when we find out. The real location of meaning.

Jon: 23:35

All right. See you, Matt.