mrb : home

The Art of Obsession: !!Con 2014

This is the video and unedited text of my keynote talk at !!Con 2014 in NYC. The slides can be found here. Transcription provided by the awesome Mirabai Knight - hire them for your next conference!


All right. Allow me to... Uh... Bet on this working. All right. No? Oh! There we go. Okay. All right. Let's see. This works. I'm pretty loud, so... What's that? For the screen? That's exciting for me. Is it accurate? Cool. I have bad hearing, and so... That's really cool to see that. All right. Anyway, I'll get started. So... This is a photograph of a really special person who I'd like to introduce you to today. How many people know who this person is? Okay. A few of you? Cool. So this is Harry Smith. I first saw this photograph when I was a teenager. About 14 or 15 years old. And I was at my friend Stinky's house. He's my best friend. His name is Matt, but I've called him Stinky his whole life. And the thing about this photograph is that it was in a book of photographs by Allen Ginsberg, who was my favorite poet at the time. I fancied myself a bit of a poet when I was a teenager, as many of you did and I'm sure still do. So it was a really interesting photograph. I had no idea who this individual was. And the caption is really funny. It says Harry Smith, painter, archivist, anthropologist, filmmaker, hermetic alchemist, transforming milk into milk. So I had to know -- who was he? I did a bit more digging. Found a photograph of him when he was younger. He had some interesting things going on earlier in his life as well. You can tell from that smile he has a lot of knowledge in his head. So it took me a long time to figure out information about him, because at that time, it was just a little bit more challenging to figure out not only basic facts about an individual, but to get your hands on the media was particularly challenging. So it turns out that he was all those things that Ginsberg said he was and a lot more. He was an anthropologist by trade, a linguist, a translator, a painter, and perhaps most of all, a collector, a very obsessive collector. I'm going to read you a quote about him. This quote and a lot of the images in this presentation are in part due to the good work of the Getty Research Institute, part of the Getty Museum in Los Angeles. So they preserved a lot of photographs about him and did research about him and read about him. You're all reading the quote on the screen. Harry Smith liked to look for keys to the universal patterns that shape our cultures and the hidden realms of the human unconscious. He compared patterns in Native American music with the eccentric rhythms of jazz, the patterns in Seminole patchwork with Ukrainian Easter eggs, the (on screen)... Sounds pretty cool, right? He was a collector. So I'm going to show you... I'm not going to tell you so much anymore. I'm going to show you the things he collected. Tell you a little bit about them, but show you too. And perhaps most famously, he's a Grammy award winning ethnomusicologist. He collected sounds and also recorded sound himself. Some of the recorded sounds he collected he presented in this 1952 epic release three double albums with booklets, called the anthology of American folk music, from the mid20s to the early 30 s, and released them -- songs, social music, and ballads. He released in 1952 after collecting it for a long time, and you can imagine that in the '40s and '50s, collecting records wasn't really a thing. These were not artifacts that people cared about. No one thought you would want to keep them around or had any idea that this music would be lost, along with the media. So the fact that he decided to collect and document them in that anthropological way was important, because he ended up influencing quite a lot of people by releasing this music that would otherwise have been lost to history. Including Joan Baez, a well known folk singer in these parts named Bob Dylan, who was not only influenced by this music. He covered some of these songs and used a lot of them in this music. And it influenced a personal favorite musician of mine, less well known, but equally brilliant. Named John Fahey. An American guitar player who took Smith's direction a little bit even further. And here's another interesting story for a musical project he worked on. In 1954, Smith was held for a week in the Anadarko, Oklahoma jail, on suspicion of stealing guns, and there he met several Kiowa Indians who introduced him (on screen) he returned to record them, because he heard the music and liked it, and he decided to set up his equipment in singers' homes or his hotel room, rather than at the actual ceremony. Because he wanted to get commentary. And he was worried about mixing taking peyote and faithfully recording these rituals, which you can imagine might be challenging. Here's the picture of that. It's called the Kiowa peyote meeting and it shows a range of his interests. He published it on the same record label 20 years after he published the anthology of American folk music, and it shows the range of his interests and the rigor of how he was capable of applying his means of collecting to a variety of different genres of music. But he also collected a lot of other things, like eggshells on toilet paper rolls. You can't really see that. There's not a lot of light. When you see the slides later -- it's literally a box full of eggshells on toilet paper rolls. He collected gourds. These are the boxes from the Getty Museum collection of his collections. Tarot decks, he was into the occult and esoteric knowledge, and one of his other collections was paper airplanes. He collected so many paper airplanes that he donated them to the Smithsonian Air and Space Museum, and they had them there for a long time. Now they're at the Getty. But the thing about the way he collected them was fascinating. He collected for years and years found paper airplanes. He lived in New York and he would walk all over the place and ended up finding a lot of paper airplanes. You never know what you're going to find if you're not looking for it. So I'm going to show you some of them. They hold fascinating cultural meaning, because as you see, the airplanes that he finds reflect components of the culture that they were found in. Like this one from 1967... That's on a flyer entitled many smokes and spring seasonings, published by the organization to end the war in Vietnam. Or this from the cafe Agogo. Published by the American Folkways in 1967. Demonstrate against the war, the Vietnam peace parade committee. Here's a paper airplane made from that flyer, a flyer from the Empire State Building. People used to like to fold them up on the top of the Empire State Building and throw them off and he would find them somewhere. A 1978 menu from Max's Kansas City, inscribed with the word Excelsior over and over again, and here's a page from a child's connect the dots book. Whatever you want to call it. So besides paper airplanes, he loved string figures. He was really interested in string figures. Here's an awesome picture of him with some string figures. It's really... You know, very photogenic individual. So here's him, recreating a string figure of some kind. Here's a sketch that he made. Of a string figure from an unpublished manuscript on string figures that was left behind when he died. So he collected string figures, which is an interesting thing to collect. How do you collect them? Right? They're an ephemeral thing, but they're a reproducible form, because you can compose them. You can start with a string of a known dimension and perform certain operations on them and then end up with that string. Right? And so he did that. And so here are some photographs of the ones that he placed, and he would see someone making it, write it down, reproduce it, and photograph it.

And they look really cool. So much so that a gallery in 2012 in Brooklyn did a recreation of some of the facsimiles of his books. So they're really cool. So besides all the other things he collected -- images. He was a really famous painter, but also a really well known experimental filmmaker. He wound up in the Bay Area in the 50s, becoming a very well known avant garde filmmaker. Made perhaps one of the most beautiful and difficult abstract films of all time, called film number 18, also called mahogany. A very challenging film to project, a very challenging film to watch, very long, very abstract, but shows that he was a collector and a composer. So he used a lot of collected, found imagery. He would find film reels, he would record everyday life, and he would compose it into these grand collages, that were meant to invoke, you know, this idea of universal consciousness that he was so desperately seeking.

Here's a couple more stills. This one is more indicative of his abstract work. So he also collected... A lot of other things. Books. Pop-up books, beaded costumes, things shaped like other things. Including spoons shaped like ducts, banks shaped like apples, and anything shaped like a hamburger. So I think we can all agree that Harry Smith was a master of obsession. Which is why I jumped right into telling you this story. This talk is called the art of obsession. My name is Michael Bernstein. This is !!Con. I'm really excited to be here. Thanks for having me. It's an honor. This is the requisite title slide. This is my first keynote talk, if you can't tell, and that's enough about me. Let's get back to the subject. So why did I tell you his story? Because I'm interested in the following questions and I wanted an excuse to ask them to you. Because you're such a room full of beautiful and intelligent people. So what can we learn from Smith's techniques? That's one thing that I wanted to ask you. He found satisfaction in knowledge, in the intersection between his interests. Right? So it wasn't that he liked music, and he liked film and he liked paper airplanes and he liked string figures. Right? He liked connections between things. He liked knowledge. And he really... He liked life. Life wasn't always so kind to him, but he really liked it, and he wanted to experience it to the fullest. So paper airplanes were interesting, because of the stories they told as a whole. Each of them is an interesting artifact. Right? They're cool, but if I said he collected a paper airplane, that would hardly be as fascinating as the fact that he collected so many over such a long period of time, in so many different places. And he actually sat and thought about what that meant. Why did people make them? What did finding them mean? What can you... What could you read from a map of reconstructed locations of found paper airplanes over space and time?

So paintings, songs, things that look like other things. They all communicate ideas. They're more than the substance that composes them. So that's one question. Another one is: how can we master this art of obsession? And why do we want to? It's summed up to me by this quote that we saw from the Getty. I love the writing that they did about his work. So I'm going to read the same quote twice. Harry Smith liked to look for keys to the universal patterns that shape our cultures and the hidden realms of the human unconscious. Keys to universal patterns. This is the -- I really like this phrase the best. From the work about him. So we can master the art of obsession by looking for these keys. That's my thesis. And the universal patterns are -- since universal patterns are composed of connections, I have a connection that I'd like to share. Which is that since string figures and knots were some of Smith's favorite things, that's a really interesting thing to me. He believed that they held some secrets to universal understanding. Knot tying can be found in many cultures, but not others, and there was a connection in Smith's mind between -- you know, how cultures and societies communicated with each other through written and spoken language and what kind of knot tying culture they had. He believed that they worked similarly to how the brain understood written numbers and letters. So Smith was really into knots. And also computer scientists and mathematicians are also really into knots. This was a figure from a book about ternary writing, and this picture that's shown here has parallels to parallel execution and term rewriting. So you can look at this knot untying as a complex term that can be normalized to a specific value. And this connection is really deep and interesting to me. And it made me think that...knots are programs. And they signal -- what is interesting about computation. Because computation is more than how computers work. Computation -- our understanding of computation influences how we understand nature the world around us, influences how we live and communicate, and it's one of the deepest human achievements, intellectually, that we've ever achieved. The achievements. You achieve them. That's how you get them. That's how they become achievements. So since knots are programs, that leads me to believe that Harry Smith would have loved computer science. It has universal application. It gets to the heart of so many human problems. Even though we don't always talk about it in the most humane terms. It has its own codes, its own languages, it's very esoteric in its own way, and it can be applied in a variety of contexts that are not immediately obvious by looking at it on the surface. And so if Harry Smith had some advice for us, he would have told us to get obsessed. If we want to learn more about computer science and understand it better, we should probably get really into it, and get obsessed in the way he did, by looking at the connections. Looking for connections between things. Because obsession is crucial to interdisciplinary thinking. If you want to understand one thing deeply, and you have your eyes open, and are available to -- making connections between the thing that you're studying and the world around you, then you can get a lot further, and Smith demonstrated this, in my mind. And the reason we want to think interdisciplinarily is because interdisciplinary thinking is crucial to advancement. The biggest, most radical achievements within the world of computer science, in my opinion, are those that have made connections to other disciplines and have leveraged other lessons from those other disciplines, and people that are aware of the world around them can make big, radical changes, in the way that we think about things, even as deep as computation. So there's so many possibilities, when it comes to computers and computer science. There's a lot of starting points for exploration. Things that you can get obsessed with. It's very fractal in nature, as it's often said. You keep studying and studying and studying, and you look at what you're studying and you realize... Wow, this looks a lot like what I think I was studying three months ago, and I'm studying the same thing, but it's somehow different. How am I solving the same problem over and over again? You're in a knot or something like that. If I was a skilled speaker, I would have a knot analogy now too, but I don't. So what kind of things with computers should we get obsessed with? I really like compilers. They're my favorite type of program lately. I think that compilers are... Compilers are the mahogany, film number 18, of computer science. You know? They're so amazing. They do everything. We... People spend a lot of time thinking about them, talking about them. If you want to teach me about how to write a compiler, you can... I'd love to chat with you after this talk. I like type systems a lot lately too. They get to the heart of how we express programs. There are these running programs that tell us information about our programs. They tell us what we can and what we cannot express. And they're there, whether we want them to be or not. It's sort of this funny thing, that you can program in one so-called paradigm and be influenced by this idea that you can choose to acknowledge it or not. And it's still sort of, to a certain extent -- influences what you're capable of expressing, in your language of choice. I'm absolutely fascinated by them, and I think they're one of the deepest... They're clearly one of the deepest connections that we have between computer science and mathematics and logic, for one thing. So I think that Harry Smith probably would have gotten a really big kick out of the idea that programs are really connected to this 1920s logic technology. He probably would have said told you so, or something like that. Static analysis is cool. It's also my day job. But programs that analyze other programs are really interesting to think about. Right? Talk about the connections between things. So I want to write my... I'm going to write a program in one language that tells you information about the program that you wrote in another language, because in the language that I'm writing in, it's good for telling you things about your programs, but I wouldn't want to write the kinds of programs that you write in the language that you write them in. Or something. So that's a lot. That's a big bag, that I would be careful before I open that one up, if I were you. Logic programming is really cool. You can solve problems that you didn't know you have.

And in understanding and learning how to use logic programming, you can shatter your narrow paradigm world views. I have a lot of exclamation points in my notes. Just my... Title was lacking one, so I felt like I needed to make up for them. I'm not speaking in a very exclamatory way right now, but they're there. I spoke a lot and thought a lot about distributed systems and they're also cool and they're also one of those fractal problems. The same problems that you have quote-unquote locally appear in a distributed way in a really different and interesting -- the properties slightly change, in other words. Things look different when you see them from afar, and then you stare at them for a while, and they kind of look the same, but not quite. That's what abstraction is, I guess. So you should study them. The world is already distributed. It's time to catch up. Exclamation point. Programming language theory. It encompasses a lot of other things. Programming language theorists and type theorists have this cool ability to somehow talk about philosophy. But now you're in the context of computer science. Right? So when you're thinking about how we express the... How we interact with... What is our interface to computation, and to computers? It's through these languages. Right? So they're clearly very powerful. And they... And it's well known that the language that you use influences what you can express. Right? That's sort of a given. But it also influences how you think about computation in general. So if you speak enough in a certain language and you think you can communicate certain things, you write enough programs in a certain programming language, and it influences how you think for a long time. You know, I learned very... I've learned very uninteresting things when I was learning how computer programming worked. I wish that I had learned other paradigms first, because then I would not have such a broken brain right now. But anyway... Yeah. How you learn really influences what you think. And programming language theory encompasses a lot of that. So how do you learn all of those things?

How do you get obsessed, and stay obsessed? How do you understand these subjects in a way that will lead you to find the types of connections that Smith made and was so good at making? So here's a little bit of advice on my take on that. This is my impractical guide to the art of obsession. So the first piece of advice is to follow the links. So find a resource. Look for the influences and connections between these resources and disciplines. Here's a metaphor for that. Your favorite band, growing up, liked music that you thought was terrible. Right? So I have a really funny story. How many people like the band Black Flag? You guys like punk rock music? Hardcore music? You know who they are, at least? If you don't, that's okay. They're a really famous band, and they were really tough. They were from California. And they thought that they were really cool. They were. They were really cool. The lead guitarist had a dirty secret, though. He loved the Grateful Dead. They were his favorite band. He wore Grateful Dead T-shirts and people thought that he was kidding. So these punks would beat up hippies and would be wearing Black Flag T-shirts while they were doing so. So the idea there is that... It's not enough to learn what your influencers teach you. Right? You need to look back at what influenced them. In order to understand how they arrived at that conclusion. That's a fundamental component to understanding how the world works. How computation works. And in order to make those connections, you have to be willing to step further than what's obvious. And out of your comfort zone.

Because I love the Grateful Dead and Black Flag a lot. They're both really great. Yeah. Cool. I'm glad I got to tell that story. So open your eyes. Be open to connections to your interests and unlikely places. This kind of goes with that. Like, not only... You know, when you're doing something completely different, be open to the idea that you might learn something about what you're not expecting to learn about. I've gotten a lot out of that, and I think Smith did too. He heard things in -- you know, he saw patterns in Ukrainian Easter eggs that were also present in... You know, whatever other beaded things that he was studying. So... I missed it from the quote, but you get my idea, I hope. Pursue discomfort.

Regularly consult texts and concepts that are too hard for you. You know, pick it up. Don't be afraid. You know, if it's a heavy book, the worst that will happen is that you'll drop it on your foot. And it won't get damaged, though, because it's a really big book. Your foot will also heal. Pick up that book and just look at what the shape of those things are that are too hard for you, and, like, stare at the page that's impossible for you to understand. And remember that feeling. Because then when you study, and you get past that, and you squash it, right, you know, you squash that discomfort, and you learn, then you get to feel really good. Right? You're like... Ha-ha! I remember when this looked like absolute gibberish to me, and now I'm talking about it on Twitter like I'm an authority.

So... Over time, watch them become understandable as you work hard and learn. Trust yourself, because you're awesome. Exclamation point. You know, if you don't trust yourself, and you're lucky enough to have other people that trust you, you'll probably still be okay, but if you really want to get as far as you can, you need to be the one to convince yourself that you're doing the right thing, and it's okay to do the wrong thing. You'll do many more wrong things than right things. If you follow Harry Smith's example, especially.

Contact your heroes. This is a good one. Because they want to teach and learn also. You know, they have things to learn from you. Whatever you've done in your life is definitely not the same as whatever your hero has done. Right? You think about individuals that accomplish things as somehow projections of those things that they've accomplished, but actually, they're just bags of meat. Like you are. And they want to be friends with you and talk to you, most likely. And if they don't, then that's their loss, not yours. You have as much to teach as you have to learn. And that's a really important thing to remember. People want to know what you think about the work that they've done. That's why they do it.

Whether they want to admit it or not. I've had a lot of awesome experiences, particularly in the world of academic computer science, where, as an outsider, I just write something, and then guess what? The author finds it. Because they Google themselves. You know? And they're like... I'm going to Google this paper that I wrote. And see who's... See what the pundits have to say about it. Pundit. And it feels really good. They're like -- hey, thanks for reading my paper. I'm like yo, thanks for writing that paper. That was cool. What other papers have you written? What other books should I read? That kind of thing. And forget what is practical. It's important to pursue concepts past the point of simple applicability. Don't just learn the thing and then use it and walk away from it. Once you've learned it or at least a part of it, you're in a pretty good position to -- I don't know, understand how it was made, why it was made that way, make your own decisions about how that thing should have been made, make it yourself, never use it again. Talk about it on Twitter. Whatever.

And then remember what is practical. Because it's good to ship software too, and to do things. As opposed to talking about them all the time. You can... Yeah. I'm going to stand by that. You should also do things. Don't just talk about them. It's really good. Doing things feels great. I've done it a couple times. So... Yeah. And then in the end, I guess the advice is to be at least a little bit more like Harry. Don't be too much like him. He disregarded his life and his physical form more than I would want for any of you. So... Embrace the world around you, and learn. Get obsessed, and that's it. Thank you.