Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
15e670d6e60c Kathryn Lawrence image

15e670d6e60c Kathryn Lawrence

defn
Avatar
1.1k Plays1 month ago

Twas the day before Hallowe’en when Ray and Josh sat down with Kathryn Lawrence to discuss magic, randomness, art… and Clojure. 

All the things Kathryn are on her web site https://kathrynisabelle.com/

MP3 SHA a6df5a115eeb24ef2419c67479368ec71209b27d6cfb4e4cb44ff8340464575e

Hosts

  • Josh The New One Glover
  • Ray The Old One McDermott

Guests

  • Kathryn Isabelle Lawrence

UUID v5 from SHA + People

47867164-ed77-5db4-a44a-15e670d6e60c

Transcript

Wouter's Comedic Entry and New Podcast Series

00:00:16
Speaker
Yeah, we definitely started recording when he calls me a wanker. We know that that's that's the official it's official entry point of the podcast for Wouter.
00:00:30
Speaker
Okay, so so I think, um yeah, but okay. Welcome to DeafN. We've got a new um ah new series after the episode of 100 where Vijay bowed out.
00:00:46
Speaker
and Josh came on board. Welcome back, Josh. Yeah, thanks a lot. And ah welcome back, Vijay, as a listener to the show for the first time ever, I hear. um You better support us on Patreon, Vijay. Come on now.

Meet Catherine: From Arts to Programming

00:01:02
Speaker
Yeah, and welcome our guest, Catherine. Hello, Catherine. Hi. Hello. Pleasure to be here.
00:01:10
Speaker
Yeah, amazing. Can you tell us a little bit about yourself, Catherine? Like who you are, where you come from, blah, blah, blah, all that kind of stuff. A bit like a first date, you know, that we're on the podcast with. Actually, that sounds a bit c creepy because you're a woman. I'm sorry about that. that's ah We can scratch that. Yeah, Velter will remove that. and We can cut that out and you can overdub it to say job interview. Oh my God.
00:01:38
Speaker
ah Exactly. Yeah, yeah how would how would you exactly would you describe yourself? if this is I think that's even worse to some extent. you know Tell me about a time when you were interviewed by two fucking white guys. Oh, many, many times that that has happened. Yeah, so I'm Catherine. You would not know me from anything. I am a software engineer. And I would have in the past, perhaps described myself as a creative technologist. um My background is not in computer science, ah but in the arts. So I am a completely self taught programmer. And about two years ago, I learned closure
00:02:25
Speaker
to join Pitch, which is a hopefully well-known closure startup in Berlin, Germany that uses a full closure stack to build a presentation software. And yes, I had been ah programming for the purposes of art and technology and making money. And Pitch completely converted me to being a closureista for life, I hope. Perfect. Amazing. It's absolutely perfect.
00:02:55
Speaker
And I would know you from pitch. And I'm sure some of our other listeners are also former pitchers or current pitchers. It's possible. Hey, Phil. um Shout out to the pitch fam. That's right.
00:03:12
Speaker
Actually, yeah, there were quite, I think ah Germany does account for 3% of the listenership. So, you know, that's good. I know Phil listens to this. Also like Adam, the CEO listens to the show. Wow. No pressure. Okay.

Life Changes: Moving and Career Shifts

00:03:30
Speaker
Well, unfortunately, as I shared before we started officially recording, I will be leaving pitch at the end of the year and returning to the United States of America because I did get engaged this summer. So I'm hoping to
00:03:45
Speaker
continue to spread the gospel of closure, join another closure shop stateside, and of course, wishing all of the best to the future of pitch and the success of all of my present and former colleagues. Oh my God, this is a tearjerker of an episode I can tell.
00:04:04
Speaker
Yeah, well, congratulations. First of all, you know, I mean, I think, uh, I think the good news is that, you know, you're, you've fallen in love, you're getting married. I mean, this is a fantastic journey. The fact that, uh, with closure, you mean, or, Oh, I see. Yeah. Okay. Well, maybe there's two love stories here. You know, there's the, let me tell you my, oh, okay.
00:04:24
Speaker
My fiance is actually a go maintainer. oh my god and ah though i I will be on enemy territory staring down that bug-eyed little blue freak of a mascot they have every day. Does he at least use Z-Max? Please tell me he uses Z-Max. I couldn't tell you. It's not the sort of thing that we discuss.
00:04:49
Speaker
um Well, I mean, you keep the Emax Vi thing out of your marriage and the tabs and spaces as well. you know This will be so be a happy marriage. You don't want to know. Just don't ask. You don't want to know. Exactly. Exactly. Happy wife, happy life. We don't need to discuss such controversial topics within our household. I mean, error handling is a completely different thing. know oh You don't get the go people started with error handling. They're they are sure everybody's doing it wrong.
00:05:20
Speaker
Well, so we said it's fine. you know yeah Anyway, Catherine, yes. So yeah, different loves and

Programming Journey: Languages and Challenges

00:05:28
Speaker
yeah, go. Very well. Okay. I think, you know, your capacity for love is clearly wide. So that's, ah that's phenomenal. ah So before you join pitch, what kind of things were you doing? what Were you JavaScripting? Were you Rubying? Were you doing something completely different but with your kind of productive, your your creative programming?
00:05:50
Speaker
Yes, I come from a JavaScript background in terms of work experience. The other German startup that I worked for before, which I will not name on this podcast, lest I am tempted to slander them. Was a JavaScript ah full stack, um but I being.
00:06:13
Speaker
a totally voracious learner have also dabbled in some Python for web scraping purposes and AI training, ah which I am actually totally not um biased against. I think a lot of people are perhaps biased against Python as like an unserious programming language for building applications with. But I actually first started programming in Python when I was working in the marketing department of that same unnamed startup, because we simply didn't have engineering resources to do things like import a bunch of product descriptions to a webpage for the content management team.
00:07:02
Speaker
And because we were lacking those resources in engineering and I had 300 new product descriptions to update, I said, fuck this. I will write a Python script to do it. I and so simply will just learn it myself. So most of the programming languages I have learned have been out of necessity.
00:07:25
Speaker
um And I find that they are all useful for different things, you know, totally. And we would never on deaf and never, never, never shit on a programming language. Like all technologies are wonderful. If they're wonderful for the users, except of course, for node JS, which is just fucking garbage.
00:07:44
Speaker
Not a programming language though in fairness. True. so Not a programming language. Yeah. And in, in the creative coding communities that I have been in, where people use a lot of strange things like processing and they're doing cool things with like shaders and even like lower level language stuff to make like glitch art, like really messing with files and, and codecs and things. Um, I find that, uh, there is a very,
00:08:13
Speaker
strict kind of policy of no tech bashing. When someone is showing their creative work and they say, I built this with this tool. If you say, why didn't you use this tool? This tool is way better.
00:08:28
Speaker
um it it's just considered really um rude and inappropriate. Although tech bashing can be really fun, especially when you're on a closer podcast where we have the best language to advocate for. It's the roster it's the comedy roasting, isn't it? you know That's what we're doing here. It's all in fun. But I totally agree, actually, Catherine, that's something. So I did the um Closure Bridge training. I know that's a big thing that the Berlin community used to do, Closure Bridge. And so we wanted to put one on in Stockholm and actually got a mentor from the Berlin Closure Bridge fam. And that was one of the things in the training. They were like, you know, absolutely no hating on technology, Windows, whatever, because, you know, that's you don't want to. Yeah, you get me.
00:09:22
Speaker
You don't want to yuck anybody's yum.

Creative Coding Culture and Community

00:09:26
Speaker
Exactly. You don't want to alienate people. you know when you're you're You're trying to show them something interesting rather than necessarily something better. They'll eventually arrive at the fact that it's better on their own. you know yeah yeah um Or not.
00:09:41
Speaker
you know ah Some pearls before swine. you know but Anyway, enough of the comedy roasting. Catherine, back to you. Back to you. You're the guest. We need to have you talking the vast majority of the time and not you, Josh. Shut up. I'm not ok talking, Ray. God damn it. Right. Catherine.
00:10:02
Speaker
Tell us a bit about your kind of like, what what was the, what was the feeling at first when you like were show enclosure or whatever? Cause this is a, it is a very in, in some ways a very strange language. If you're used to Python, especially or JavaScript, you know, it's, it's kind of different. So was that your initial reaction or did you kind of fall in love or what happened? You know, what was the story there?
00:10:30
Speaker
Yeah, coming from a JavaScript shop to closure, I knew I was going to have to just learn it to start the job. And Pitch was very kind. They said you they hired me knowing I did not know closure. um But they believed I had the capability to learn it. And they said, you do not have to learn it before you start this job. You you can learn it but through your work here.
00:10:55
Speaker
But I am a bit of an overachiever and a bit of a teacher's pet. So i bought I bought the book, Closure for the Brave and True, by Daniel Higgin, bought them as my first introduction to closure. Big shout out to Daniel, yeah.
00:11:13
Speaker
Yeah. And I read that cover to cover and then I built a toy project enclosure to get some hands-on experience with the REPL and the build tools. um I made actually an astrology application. So I used the Swiss ephemeris, which is like a data source that tells you basically at any given time where we ah could measure the position of the planets in relation to the latitude and longitudes of the Earth. So it's kind of a geographical mapping database. And someone has created closure bindings for the Swiss ephemeris. So there was my back end.

Astrology, Occult, and Tech: Catherine's Projects

00:11:59
Speaker
And for the front end, I used closure 2D to basically just draw the chart, the astrological wheel you may be familiar with.
00:12:09
Speaker
or or maybe our listeners are familiar with the chart of the moment. So you have all of the houses and signs and then the degree that the planet is sitting on within the house and sign, which some believe can help you interpret the current state of the universe and influences on the workings of mankind and other things on the earth. So closure 2D for the visualization is very simple.
00:12:38
Speaker
kind of chart that I was drawing. And yeah, line engine for the like build tool. And that was my first ever closure app was a little desktop astrology chart.
00:12:52
Speaker
application. And then of course, I hit the ground running at pitch. And using closure script is not so different from JavaScript. I was very familiar already with building react components, I just had to do it with a bunch of parentheses instead of a bucket. And when we converted our front end from using reagent to uix, it was even closer to my known ah experience and way of writing React hooks and things in components. so The transition actually was very easy once I got familiar with the syntax of the language and ah building React components exactly as I always had. nice yeah and so Where are you pointing him at it, Ray? I don't know what I'm supposed to say. I'm new to this whole
00:13:45
Speaker
Dethan hosting thing. yeah yeah but So, Catherine, we... I mean, Pitch is a remote first company, or or at least post-pandemic we were, and so I joined um and When was that actually? I joined in like 2022 maybe. And so like we actually didn't know each other very well at pitch other than in Slack, but then we did this really cool like offsite and we had a mini conference with tech talks and you gave a really cool talk on this askew tarot thing that you did.
00:14:23
Speaker
and um and you know random numbers and all sorts of stuff. So that was kind of the first you know big dose of Catherine I got right in the old main vein. And I was like, holy shit, this person is is somebody I need to talk to. So like can you tell us about like why was I so blown away by your talk?
00:14:45
Speaker
Well, I'm extremely flattered that you would be. um But yeah, I guess having already covered the astrology ah portion of my interests, I think the intersection of the occult arts and technology is a very rich and interesting one. And I have given some talks in more of a tech ah creative space about tricking computers into doing magic for us.
00:15:14
Speaker
because we may discuss that ah most methods of divination, if you want to predict the future, if you want to ask a question to the universe, to a higher power, to whatever you believe in, the way that you usually accomplish getting your answer is through some kind of random process. So going back as far as you know the the ancient Taoist practice of casting sticks. They have a ah method of divination to find a corresponding hexagram in the Yijing by taking a bundle of 50 sticks and sorting them in a way that produces kind of a pseudo random number. And then what that number corresponds to
00:16:05
Speaker
in the book, the hexagram shape in the Yijing will tell you, oh, it is an auspicious time to plant crops, or it is an inauspicious time to go to war. And you could interpret that, you know, for the emperor, however you wish to, but it's kind of just a vibe check. So when you flip a coin or shake a magic eight ball, what you're doing is something that has a completely more or less random output. And we can talk about the various probabilities that are inherent to the the methods of generating a random output. um But that is so rich to explore with technology. So
00:16:48
Speaker
in trying to do divination like tarot card reading or astrological interpretation or using the large language models as some kind of magic eight ball ah content producing answer producing mysterious method um has been very ah productive and interesting to me as far as ah creative technological pursuits go.
00:17:14
Speaker
Okay, so that is the only valid application of LLMs producing some bullshit that is interesting. Yeah, absolutely. lee And the way that I have used AI, ah we should not get onto an AI art tangent here. But the way that I have used AI is very similar to this ah cut up method that was used by chaos magicians in the 70s and 80s, also used by ah the poet William Burroughs, also used by David Bowie to generate some of his lyrics, where you take a bunch of words from newspapers or diaries or some text corpus that has significance to you, right, and you randomize them. And then whatever words you pull out of the hat at that moment, ah create a kind of mystical significance just because it is the random output of those
00:18:08
Speaker
things and and you can use a large language model in very much the same way because it is a more or less random depending on the temperature, right? that If you increase the temperature of what you're requesting from the model, it will be less certain that the output is actually ah statistically common.
00:18:30
Speaker
And in this case, the LLM is presumably a locally hosted LLM rather than some monstrous thing that burns children for breakfast. Yeah, I gave a talk about AI for artists where I really advocate for training your own models, which inherently have to be smaller models and curating your own interesting data sets. So some of my previous work that I did a lot of web scraping for, back to my Python scripting days, um was generating a TED talk where I scraped about 3,000 transcripts and fed them into an LLM. um The title of that talk was Yeah, the the title of that talk was memorials to our humanity and it was featured in an an online art exhibition in 2019 that actually got a cease and desist letter from the TED organization. Oh but that is the oh my god oh god. That was so cool. They said you cannot use this color red, you cannot use our logo, you are not associated with TED because it was very much satirically billed as being a TEDx event as part of the wrong Biennale.
00:19:45
Speaker
Yeah. um Also, I sorry sorry just um wanted to talk about two more interesting datasets. One of them was scraping Yahoo answers before it was shut down. That was kind of like the burning of the library of Alexandria to me personally. yeah A lot of interesting text content

Randomness and AI in Creative Work

00:20:04
Speaker
there. And I have an upcoming project where I'm planning to use text from the website Aerowid.
00:20:10
Speaker
where people write about their experiences using various drugs and basically do MK ultra on an LLM by only feeding it people's trip reports, um kind of playing with the concept of AI hallucination. Oh, this is so good. Yeah. Yeah. So there's a lot of interesting data sets out there and you can make your own. And that's part of the fun of it. Right. And that's part of the magic, like a lot of the magical texts and ah let's say grimoires, some of the the known methods of doing magic involve creating your own tools. So maybe that's how I can pivot to talking about my tarot deck, which Josh has a copy of. Totally. And just like, please, please, please, Catherine, just shower us with links so we can put stuff in the show notes because now I want to just binge your content. This is too exciting.
00:21:05
Speaker
yeah i can we stop recording i'll just go do that i just updated my personal website to have functioning links to everything so it is katherine isabelle dot com it should be available in the show notes i will send you a link i do spell katherine with a k and an r y m isabelle is spelled with the full e l l e at the end ah um and yes so
00:21:35
Speaker
Oh, I lost my train of thought. We're going to pivot to the tarot cards. Right, right. So I was involved in some glitch artist communities online. This was actually before Discord, back when we were on social Slack channels. um Before we got hip to what the kids were doing on Discord, I was in a social Slack with some glitch artists.
00:21:56
Speaker
And we were discussing magic, glitch, a lot of those people were making their own tools for generating art. And I was gifted my first tarot deck. And I was inspired to create a deck using ASCII art because that has been an absolute passion of mine as a nerd and someone who loves to add a bit of whimsy to my projects. You can never go wrong adding some ASCII sparkles or a console log that draws something very interesting. You know, if you start any project just setting up the infrastructure, the first thing you should do is make sure that it can print a cute little
00:22:39
Speaker
picture of something. totally preach yeah So I used Nano as my text editor, and let's not have any discussion about whether that was the best tool for the job. But I illustrated yeah the the full deck, i all of the cards, as an adaptation of the writer's myth-weight tarot deck. That is the one most people would be familiar with, the illustrations of the illustrations um by Pamela Coleman-Smith. You can look it up on Wikipedia. I think it it it would be the entry for tarot deck or writer weight deck. And they were all illustrated just in 20 lines of text. And they were for a command line interface that just uses ah JavaScript math dot random as its random number generator.
00:23:37
Speaker
And I found that that was more or less sufficient for mystical synchronicity. But yeah, it was originally illustrated to be used on the command line. So they're very small ASCII illustrations that are fit into only 20 lines of text. And I actually later adapted it to an IRL version where I printed the text in, of course, a lime green kind of yeah homebrew shell ah theme. What is beautiful is the theme? What is the terminal theme called? It's green on black. It's like the homebrew, not homebrew. It's the homebrew theme and in green and black on paper. So the IRL deck you can actually, you know, use as a deck of cards.
00:24:28
Speaker
And just ah last year, the end of last year or early this year, I created a Discord bot adaptation because now I am hip and cool and I'm on Discord with all my friends. And you can um use a ah regular Discord bot to ah use the draw command and it will draw a card and all the images are ah just images from the same deck, but that one uses the actual random.org API, which is slightly more legitimately random than JavaScript's math dot.random function, because random dot.org uses atmospheric noise for generating random numbers, and they actually statistically test their outputs for true
00:25:22
Speaker
are as close to true randomness as possible. So I find that the Discord bot is actually um creepily more accurate than using the command line interface. But ah both of those are free and they are linked on my website if you're interested in trying them out and you haven't been as Josh was gifted the IRL edition of the deck. It's beautiful though. can People can order a copyright from your website, I think, or they used to be able to, or maybe not.
00:25:54
Speaker
Yeah, they could. There are very few existing in the world at this point, so I am accepting orders and inquiries just via email now, and once I have enough interest to print a second edition, they will be available for easy purchase with a Stripe link on my website again, but currently the the Stripe Store is disabled.
00:26:14
Speaker
Okay, nice, nice, nice. But the funny thing about it, we're talking about like ah randomness for for the cultish thing. But of course, random to me, it's, it's peculiar how ah how the magic aspect of randomness is also the basis of all internet security. We're talking about Stripe shopping and all this kind of stuff. Yeah, everything to do with ah cryptography, everything. And you know, it's kind of It amuses me. you know I can't remember which the what the author was, and Josh will probably remind me, or you will, Catherine, who said that know as um technology increases in sort of sophistication and complexity, it ah essentially becomes magic.
00:26:58
Speaker
um and you know So that's Arthur C. Clarkway. Well, actually, the quote is, now any sufficiently advanced technology is indistinguishable from magic, obviously. Oh, key dokey. Yeah. Right. Thank you for that in line. Live correction. Fact check.
00:27:15
Speaker
yeah is and And for cryptography, I mean, let's talk about lava lamps, right? You know, that cloud flare ah encryption seed generated by a wall of beautiful lava lamps that's being watched by a webcam. I mean, that's magic, is it not? Absolutely. Absolutely. You know, and but it's it's sort of funny how how um We're in a world of like logic and order, and yet at the heart of it, at the reality of it, the fundamentals of it is randomness. you know and Yeah, so it'd be interesting to get into that a bit more deeply, I think, because I know this is something that you're interested in, Catherine.
00:27:59
Speaker
so Let's talk a little bit more about that, about how how randomness is sort of like essentially everywhere. and you know I'm a believer in RNGesus, so you know I want to spread the gospel of RNGesus, so let's let's get into it. you know Yeah, praise RN Jesus, our Lord and pseudo random savior.
00:28:21
Speaker
um I think as far as randomness as like a philosophy and as like the basis of our universe goes, we need to talk about, of course, quantum theory, much beloved by physics nerds and woo woo ah magic users and occultists alike.
00:28:42
Speaker
So this is kind of the the quantum theory ray of drawing tarot cards. I know a lot of people, if they give you a tarot reading, they will spread out the deck of cards and then they'll wave their hands over it to like feel the vibes. You don't need to do all of that. All you have to do is know that before you turn over a card in the deck, it is in a superposition where it is every card in the deck. And as soon as you flip it over and observe it,
00:29:10
Speaker
you have entered into an agreement with the universe that this is the state of the universe at that moment, that this card is visible out of every other card, you have created reality, you have co-created reality through a random process of flipping a single card. and address car park Yeah. Yeah. So that's kind of the the quantum entanglement part of it. But you're totally right, right, that we spent so much of our time as programmers trying to do things deterministically. And when we have to do something that gets random, I don't know if maybe some of our listeners or you guys have ever seen the error, ah insufficient entropy.
00:29:51
Speaker
yeah yeah yeah We try to use the sources, ah every input at our disposal, the webcam, the the temperature of the computer, mouse clicks, mouse movements, everything to try and figure out, to to get a unique output and and feed all of this kind of behavior and ambient information into something that can produce ah ah something that is unpredictable, yeah and and trying to predict unpredictability, trying to force a deterministic machine to give you a non deterministic output, you do have to feed something into it from the real world. It's really making the computer interact with a larger force of the universe, whether it's atmospheric noise, whether it's the
00:30:40
Speaker
barometric movements of lava lamps, whether it's your completely random mouse movements and ah temperature fluctuations in your machine, entropy is is necessary to do any magical process. And whether it's totally random or just pseudo random, when you actually engage with the action and turn the entropy into a deterministic result, that's how you co-create reality with the random process. Damn. Yeah, I remember actually in the old days when you generated like a long GPG key, I think it used DevURandom on Linux and it would literally tell you to like move your mouse, generate some disk activities. So I was always doing like, you know, du-h on my root file system and, you know, moving the mouse like a wild man and all of this. That was ah was fun.
00:31:37
Speaker
Yeah, and did you feel like you were doing a ritual, casting a spell, kind of yeah i mean bending the computer to your will through some kind of totally bizarre sequence of actions?
00:31:50
Speaker
So I never thought of it like that, but I i think like about 15 years ago, I read this book called Hip Tastes, and it was a little 100-page book by a sommelier trying to demystify um the whole like wine culture. And um she had a whole chapter on ritual. She was like, listen, the shape of the glass, the only difference it makes is the surface area of the wine. Everything else is bullshit. However,
00:32:19
Speaker
like double-blind tests have shown that you know using using different glasses changes the user's perception of the wine. So like she got into this idea of ritual as something that brings richness to our lives. And I thought that was like super

The Magic of Programming: Rituals and Recursion

00:32:36
Speaker
interesting. And just after reading that book, I started paying attention to other things I did in my life and like what what had that kind of ritualistic thing and how I felt about it. So anyway, sorry, tangent, but that's what we do here.
00:32:50
Speaker
Yeah, and I'm sure a lot of things that programmers do could be seen as very superstitious, whether you understand the process behind them or not. I mean, there are also real um magical or religious beliefs that people imbue their computing with. I'm thinking, for example, things like having members of the clergy come to bless your data center.
00:33:15
Speaker
i Yes, which is something you can find ah hilarious pictures of on Wikipedia, um or this um belief that if you leave a certain type of crisp packet, I think it's it's a Southeast Asian um brand of of potato chips. If you put that on top of your of your web server, it will help it keep functioning. Maybe that has something to do with ghosts, not interfering with your computer, but there's all sorts of interactions between kind of like magical ah rituals and technology that are as simple as maybe the one that most of our audience would believe and be familiar with, not deploying on a Friday.
00:33:57
Speaker
Right? Yeah, that's an unlucky day to do a deploy. You just simply wouldn't do it. Yeah, I mean, I think ah the the the concept of like magic is is everywhere, I think these days, you know, and and I mean, to be honest, I think once you're programming, you you realize that um Yeah, you can you can try and tell the computer what to do, but it's pretty rare that it actually does it first time. So, you know, you're always incanting some kind of God or some kind of superpower. And what do we call computer processes that we spawn to
00:34:38
Speaker
deliver our will unto other programs, they're literally demons. They're demons. Absolutely. And we have zombies, right? If the parent dies and the child lives on, it's a zombie process. Yeah. Yep.
00:34:50
Speaker
Yeah, I will say that when I have discussed ah these concepts with other engineers, I do have the utmost respect for people who don't want to engage on the like, occultic and magical level with these technologies. I did have a conversation with a colleague recently who I was discussing all of these super exciting magical ideas with who said that um due to his personal religion he doesn't fuck with demons and I said you know what I respect that if you don't fuck with demons maybe you know a little bit more about them than I do because you have a healthy fear and respect and I'm over here dabbling over here dabbling in the dark arts and trying to bend entropy to my will who knows what's going to happen because of that so
00:35:42
Speaker
It's not even the skeptical viewpoint. It's the viewpoint of this is dangerous. Wow. Okay. Yeah. What? So running a unique system is dangerous. is that Is that basically what we're saying? Yeah. If you try and get too deep on the level of demons and stuff, absolutely.
00:36:07
Speaker
Oh man, there's so many thoughts popping off in my head now, but like maybe we should steer somewhere back into closure territory somewhat, um and unless we want to explore the idea of you know randomness a little more. Because you did threaten that you could talk about this for five to six hours, which is about the length of a normal definite episode anyway, Ray, isn't it?
00:36:31
Speaker
Yeah, without any editing. Yeah, for sure. You know, you said it down to two or three hours, but, you know, you have umma the raw cuts. Let's talk a little bit about fractals and that will bring us back towards closure, right? Because there is a lot of overlap in kind of the, again, scientifically minded and and physics and natural sciences community, right? About ah the structure of our known universe as
00:37:02
Speaker
fractal as holographic as recursive, right? And I think that thinking recursively using a language like a lis also can help you kind of tap into this kind of divine fractal nature of the universe. I remember there was a talk at the Heart of Closure conference called Standing on the Shoulders of Giants.
00:37:28
Speaker
I had discussed the history of the study of recursion and particularly this Hungarian professor um named Rosa Peters, I believe, yeah who was a major contributor to the theory of um proving recursion mathematically, like the the mathematical proofs that they were inventing in like the 20s and the 30s. And the paradox that that talk opened with whether the set of all sets that do not contain themselves and contain itself. I mean, this is basically the same question as could God create a rock so big he couldn't lift it, right? yeah And one of the main takeaways from from thinking about this study of recursive algorithms is that we we know that we can mathematically prove that some things are unprovable.
00:38:26
Speaker
And that's okay. That is perfectly fine for me personally as a justification for the existence of magic and the unknown. But going back to fractals and recursive thinking, it really gets you deep into those rabbit holes of existence, non-existence, infinity? Oh no. Yeah. I mean, yeah we we didn't mention, but it is the day before Halloween that we're recording this, so it's good that we're getting into really scary shit like, um you know do we exist? Is that the question though? Or is it, is it Emacs or some other shit, Josh? and it's like
00:39:09
Speaker
Well, I use Emacs, therefore I exist. I don't know how to say that in in Latin. Maybe I should look that up. Vijay will know. Yeah, yeah, yeah, yeah. But ah yeah, so I think the whole notion of ah recursion is ah something which, bizarrely, I think, you know you You can kind of learn it as ah as a five or six-year-old, and yet it is quite scary. I mean i remember like when I was programming in other, lesser languages, um that ah that if you saw a pull request or a kind of code change with recursion in it, everyone would like, oh, whoa, whoa, whoa, whoa, whoa, whoa, stop, stop, stop, stop, stop.
00:39:50
Speaker
We can do this without recursion. We can do this without recursion. Yeah, but it takes you 50 lines where this is three lines and it's beautiful, you know? Ah, but it could all go wrong. Yeah, yeah but it but it won, you know? We've never seen a for loop go wrong, right? There are no off by one errors ever. Yeah, but it's but it's kind of funny how there is a sort of, there is a suspicion, there is this there is a sort of, um,
00:40:16
Speaker
A fear of recursion in the programming world. I want to say that. I mean, it's not about, you know, necessary vehicle, but it is. There is a fear of. um of it's a bit of it's like, um, not just the sort of, uh, the stack rates, but this fear of. Out being things being out of control or running away. Um, yeah, it's like that scene in Fantasia, right? Where Mickey mouse is the little wizard and chanting the brooms and then it just keeps going. yeah
00:40:48
Speaker
Yeah or things just keep on spawning and spawning and spawning you know i'm and so there's this is fear that even a very well so crafted bit of recursion has a problem and i know i occasionally even enclosure i write a thing called loop record or record function you know and.
00:41:10
Speaker
A lot of closure ins will say i don't do that you know you should be using reduce map filter all these sort of things but a loop and recur they have their place you know yeah. and I want to be in that place where they where they can exist where everyone's free to loop and recur.
00:41:27
Speaker
Yeah, I also remember this fear of recursion when using other languages, but I will say that using Clojure has fundamentally altered my brain. Because as I now study, again, ah data structures and algorithms to brush up on my lead code for an impending job search in the Bay Area. I find myself tending towards and preferring the recursive version of a search algorithm because it just makes sense after you've been dealing with and sequences and nothing but sequences. Yeah. I also think some somehow, I mean, and again, maybe this is, I don't know if we're going to jump off on this one, but when you recur, there is a sort of feeling, that there's a sort of tendency towards immutability.
00:42:17
Speaker
um Where is if you're if you're kind of looking over things then you tend towards mutability you tend towards like in place changes and so when you recur you tend towards. Immutability i don't know if that's if that's a vibe or if it's um what what are your feelings about that.
00:42:35
Speaker
I don't have personal feelings about it. I will say it does inspire me to think about the different kind of, um of course, with closure being ah preferring immutable data structures, it makes sense why we would also prefer recursive recursive versions of algorithms. But Josh, do you have any thoughts?
00:42:59
Speaker
ah No, the reason I mentioned is because job interviews, you know, you mentioned the job interviews and a lot of people were when asking of randomness yeah when they when they ask you to, ah when they ask you to, I don't know, to to do some sort of beatery or whatever, then the The natural way of doing it is just recursive or recursing down through the tree, you know do a ah essentially a kind of walk. But if you're in a sort of like of traditional non-recursive kind of environment, then you will do do updates as you go along. and you will you will keep everything like you You will optimize immediately. You'll optimize for, OK, just keep ah keep this space.
00:43:43
Speaker
very tightly defined. And that dinner know that that's ah a bit of a difference, I feel like. you know At a certain scale, that might be true. It might be faster to do it. But again, a lot of it is to do with the way that comp computers are structured or the way that way that we kind of, um a lot of these algorithms and data structures and the way that systems are made come from the 1960s, 1970s, not necessarily from the mathematics of the 20s and 30s,
00:44:11
Speaker
which are more fundamental.

AI Concerns and Community Experiences

00:44:13
Speaker
um you know They tend to be constrained by oh the price of memory, the price of computer hardware, essentially. And we're seeing in this sort of but the last 10 years that um weather that we can revisit some of those assumptions. And I think we're going to find the same thing with algorithms and ah data structures as well, that over the next 10, 20 years, as as we become more liberal, let's say, with our ability to acquire hardware, then maybe some of these restraints will be gone. I'm i'm on a rant now. I'll stop. Okay. Catherine, what do you think? It just reminds me of of also some research that I did into ah neural network and the algorithms that we're using for for creating Bayesian models and things like that, where we're always looping through the nodes and
00:45:12
Speaker
Yeah, I think with the amount of computational power, time and energy that AI algorithms take, right, those would be unthinkable years ago, and we're only scaling up the infrastructure to support them and make bigger and bigger models.
00:45:32
Speaker
And I don't know if that's necessarily a great thing for the world in general, that the power is going unchecked, but it is very computationally interesting to see what we can do when we have the endless power to enumerate all possible universes and all possible paths and all possible notes. I mean, we get into Mr. Strange territory right where all known universes could be known to us and That kind of ties it back into the magical weirdness. I'm trying to bring to the pod, but It could be true who knows I think that is an AI future that we could be somewhat optimistic about and
00:46:18
Speaker
Yeah, I mean, i didn't want I definitely didn't want to motivate large scale data centers by that comment. So let's just be clear about that. you know
00:46:31
Speaker
Shout out to you know a friend of the show, another podcast, the Data Vampires series, which we should drop on just ah to talk more about data centers and what LOMs are doing to us all. But but yeah, I agree, Catherine. they're like This stuff is super fun to think about. but For me, I just love thinking about this shit, and you're excited, so say something. I would also just like to put on the record, I am not super excited for all of the water and matter in my body to be burned up and turned into paperclips personally.
00:47:07
Speaker
If I am to become a paperclip, you know, I hope that I will be a good one and that my matter will be enjoyed by the rest of the universe as it is slowly turned into paperclips as well. We must appease the Basilisk, right?
00:47:25
Speaker
I see you're making a podcast. Can I help with that? but was just That was a perfect clip joke there. Okay. I know we got it. Thanks. Thanks. All right. So we have been going for a while, not as long as usual. Cause unfortunately like I have to go and do a thing, you know, spoiler of, of the show. back It was better when VJ was this guy. god ah Anyway, sorry, sorry. ah So what I was going to say is that we ran into each other, all three of us at Heart of Closure, which you already alluded to, Catherine, what, about a month ago as of recording? And um I just like really wanted to get your thoughts about the conference um because I know I had a great time, Ray, whatever, almost got me killed, but that's another story. But what about you, Catherine?
00:48:21
Speaker
I loved it. I had so much fun. I have nothing but rave reviews and love in my heart for the closure community. It was so inspiring. It was so refreshing. Everyone was so kind and smart and interesting. I feel like I made 100 new friends in the span of of two days. And yeah, i I just loved it. It was a great work bye by the Heart of Closure organizers. Yeah. And you mentioned Daniel's talk on the the shoulders that are standing on the shoulders of giants. um what What other talks kind of stuck in your memory? Any other like, you know, let's reminisce here for a few minutes before we close it out.
00:49:09
Speaker
Yeah, of a pop quiz, which talks is I actually paying attention to. Now, Shoulders of Giants really made an impression on me. I also, I have never found, um I know this makes me kind of a ah lame um No, I won't disparage myself. People who are into compilers are weird, and I shouldn't aspire to be one of those people. But as someone who has never appreciated the buter ah the beauty of working with compilers, I really enjoyed enjoyed the talk.
00:49:42
Speaker
Let me just do that a whole sentence again. As someone who has never personally been interested in compilers, the talk by the um maintainer and I think creator of Jank about adding functions to Jank programming down to the compiler level actually made me interested in how that works at that level for the first time.
00:50:08
Speaker
And I will say that the closing, of course, the opening keynote was amazing, but you already had Lou on and everyone gushed about how amazing that was from the very first hour of the conference. So I don't need to mention that again. That was amazing. Lou's scrappy fiddles talk about what it means to be open. yeah um And the closing talk actually by Eric Normand about the wonders of abstraction. That was really enjoyable to me as someone who likes to go all the way back to the Stone Age to figure out why we're doing what we do with computers today. I really enjoyed that that talk went from the history of stone tools all the way to building via
00:50:55
Speaker
the logic gates of computer systems. That was super fascinating. and And I love that he threw it all the way back to the Stone Age with that. Yeah. Spoiler alert, we're going to have Eric on soon. so thats yeah Oh, no. Spoiler for the wonders of abstraction. But I believe you can already watch that talk online on YouTube, right? Of course, yeah. Part of closure talks are up.
00:51:16
Speaker
Yeah, but don't watch it before the death in episode. Come on. Actually his episode won't be out next, Catherine. So you you are um going to be a hard act for him to follow, but you know, he's going to have to step up. I'm just teeing you up for a segue right into the next episode. so That's the magic of podcasting.
00:51:40
Speaker
Oh, Ray, come on. oh yeah but I guess on that, like we had better just go with that terrible pun. I mean, it's only going to go downhill from here. So, um, Catherine, it was great to talk to you today. I, um, yeah, I'm really bummed that we didn't work together for long enough at pitch for us to have more interesting conversations about this stuff, but thanks at least for coming on the podcast and and telling us fun stuff.
00:52:11
Speaker
Thank you for having me. I've really enjoyed ranting about art and randomness and divination and occultism and computers. And if there's if there's only one thing that I that I would want to plug, um it is the concept of believing in magic, even though you are a person who does things with computers. You can have both. You can be a programmer, believe in deterministic systems and also believe in the great unknown and the unknowable forces of the universe and harness their powers to your dark intention.
00:52:51
Speaker
The dark intentions bit that's slightly worrying to me there, but ah do your listeners. It is almost Halloween, so I thought I would get a little spooky with it. Thanks a lot, Catherine. Best of luck with your return to the US s and you know all the activities and the job hunts and the marriage and those kind of stuff. and No doubt we'll hear from you again, because ah you're ah clearly an absolutely fabulous person doing great stuff and very creative. And I've really enjoyed this conversation as well. So thank you. Thank you very, very much. Thank you so much. I hope to continue to also be a part of this great closure community. You guys are doing great work as well. All right. On that note, we're out.
00:53:41
Speaker
Thank you for listening to this episode of DeafN and the awesome vegetarian music on the track is Melon Hamburger by Pizzeri and the show's audio is mixed by Wouter Dalert. I'm pretty sure I butchered his name. Umm maybe you should insert your own name here Dalert.
00:53:58
Speaker
but If you'd like to support us, please do check out our Patreon page and you can show your appreciation to all the hard work or the lack of hard work that we're doing. And um you can also catch up with either Ray with me for some unexplainable reason ah you want to interact with us, then do check us out on Slack, Closureion Slack or Closureverse or on Zulep or just at us at Deafened Podcast on Twitter.
00:54:27
Speaker
Enjoy your day and see you in the next episode.