Everest Pipkin: Hey. I’m Everest Pipkin. This hap­pened at my last talk which I gave, and I was real­ly into it because I was talk­ing about Byzantium and I had this beau­ti­ful glow­ing back­ground behind me and it was real­ly fab­u­lous. So, I don’t know, maybe this is just the way it is.

But, yeah. I’m an iter­a­tive bot-maker, among oth­er things. I’m going to talk about some slight­ly embar­rass­ing aspects of my work. And if you don’t know what I do, it would be amaz­ing per­son­al­ly if you would look up the rest of my work. My web site is ever​est​-pip​kin​.com. But I thought it was kind of impor­tant to maybe expose the veil and talk about some of those oth­er aspects of what I did to be here. So don’t remem­ber me as just this, but it’s kind of great.

I want­ed to talk about how bots and specif­i­cal­ly Twitter bots are kind of func­tion­ing soci­etal­ly today. I thought we could start with the idea of folk art. There’s a lot of ways one could define folk artis­tic prac­tice, but to cut to some of my gen­er­al­iza­tions, it gen­er­al­ly exists out­side of fine art cir­cles, or at least oper­ates inde­pen­dent­ly (some­times will­ing­ly) from cur­rent main­stream trends. It’s often uncom­mer­cial or exists in hand with per­son­al sus­te­nance economies. This is to say that folk prac­tice may sup­port the artist but rarely sup­ports an insti­tu­tion at the same time. And it tends to uti­lize afford­able and acces­si­ble mate­ri­als which replace expen­sive art mate­ri­als as needed.

All of these kinds of prac­tices man­i­fest dig­i­tal­ly as well as they do in phys­i­cal space. There are many many many dig­i­tal cre­ators who are oper­at­ing out­side of the insti­tu­tion. And one of the bet­ter aspects of the Internet is the avail­abil­i­ty of Arte Povera dig­i­tal mate­ri­als. Which is to say there are many open APIs, strange data solu­tions, and huge cor­po­ra sources in the pub­lic domain.

Of course, dig­i­tal folk art is not a new phe­nom­e­non. People have been work­ing cre­ative­ly out­side of the gallery space as long as the Internet has exist­ed. In some ways the two go hand in hand. The ear­ly Internet was very much a place for enthu­si­asts, hob­by­ists, tin­ker­ers, and much of the con­tent of the ear­ly Internet was sim­i­lar­ly off-beat. It was not con­sid­ered a place to be tak­en seri­ous­ly, and so seri­ous work tend­ed to avoid it. And even as Internet com­mu­ni­ties grew and com­mer­cial­ized as they did, this men­tal­i­ty’s flour­ished with the vast avail­abil­i­ty of online spaces.

So you might want to con­sid­er fan prac­tices, includ­ing fan fic­tion and role­play. I was going to show this real­ly embar­rass­ing web site I made when I was thir­teen years old on Avid Gamers, this fan­ta­sy role-playing site. But you can just imag­ine it. It was ter­ri­ble. But impor­tant, right? And GeoCities pages, which sort of exist some­where between a col­lage and a self por­trait. Or web comics. Or per­haps a con­sid­er­able amount of DeviantArt.

And I’m not pok­ing fun at these com­mu­ni­ties. I spent my child­hood and my ear­ly teen years in them, and I’ve nev­er worked so hard or so pas­sion­ate­ly. And I’m end­less­ly thank­ful for the lessons I learned there. And I’m still impressed with the work that comes out of them, in all its sil­ly and won­der­ful Internet glory.

And bots are real­ly not so dif­fer­ent. So for exam­ple, very few of them are made for the gallery. Very few of them are made for com­mer­cial rea­sons. As a group, they have their own set of zany and beau­ti­ful strange aes­thet­ics. The com­mu­ni­ty that sup­ports them and pro­duces them is rel­a­tive­ly small and hand­made. Often, bot-makers are not trained in the arts. And even when mak­ers of bots do come from tra­di­tion­al art spaces, like me, or even when they’re express­ly made for the gallery, they remain fair­ly egal­i­tar­i­an. Which is to say that although bots kind of assume that their view­er will have access to the Internet (which is of course priv­i­lege), they’re also avail­able every­where, and view­able in an orig­i­nal for­mat. So unlike a gallery show that requires a vis­i­tor to go to the phys­i­cal loca­tion and see the thing to have a phys­i­cal expe­ri­ence with the work, bots are orig­i­nal every­where when you’re with them in space. They live with us.

So this is excit­ing for a lot of rea­sons. For one, it’s polit­i­cal. Making work inside of the cor­po­ra­tized Internet, but devel­op­ing shared toolsets with a com­mu­ni­ty of oth­er mak­ers for the joy of it is push­ing back. And it’s weird, right? The type of projects that come out of this space are very dif­fer­ent than any oth­er space I’ve ever seen. We should be proud of that. And its ours. The space serves us because we built the thing.

But I also want to talk about the inverse of the sit­u­a­tion, like the rogue or folk com­put­er sci­en­tist. Which is to say that rela­tion­ship flows both ways. I’m inter­est­ed in what hap­pens when artists who are used to being artists decide that the best place for a work is with­in a space that seems to require an entire­ly dif­fer­ent method of con­struc­tion. And of course, there’s no harsh line between forms, and plen­ty of peo­ple exist both as highly-proficient work­ing artists and excep­tion­al­ly skilled pro­gram­mers. Tons of them, right? But I’m not talk­ing so much about the skill or even back­ground. Instead of I’m inter­est­ed in men­tal­i­ty. [The] goals of each prac­tice, and how that comes to bear on result­ing work.

So often, the under­stand­ing of how code should work is that it should be invis­i­ble, right? Good code should­n’t make itself known. It’s meant to pro­vide this sort of seam­less dig­i­tal expe­ri­ence, one that is not jar­ring or sud­den. And okay, this makes sense, since yes usable design is a good busi­ness plan. Folks pro­gram things like auto­mat­ed vehi­cles and kitchen appli­ances. They prob­a­bly should­n’t kill you acci­den­tal­ly because of an art­ful­ly bro­ken inter­face. It’d be bad. But it actu­al­ly makes lit­tle sense as a base­line in the arts. 

So, side­note. I start­ed mak­ing bots with no idea of what I was doing at all. I’d nev­er used command-line tools. I’d nev­er used node. I’d sort of stag­gered through a few Javascript tuto­ri­als and made some web sites, but I had no struc­tur­al under­stand­ing of how to write or han­dle code at all. And I was liv­ing alone in Rome, Minnesota, and I was like, Alright, I’m going to do this thing.” 

So I was going to show the first pass of the code for tiny star fields, which I can just describe. You’ll get the idea. It’s two hun­dred lines long for an incred­i­bly sim­ple bot. I did­n’t know that you could just type char­ac­ters into Javascript and have it use strings. So it goes through an out­side text file and pulls an indi­vid­ual line into the bot, which it then pop­u­lates a sin­gle char­ac­ter with. And it does all of these not in a for loop, because I did not know what for loops [were]. So, it did this like forty-four times, essen­tial­ly. It would pull a sin­gle line of a Unicode char­ac­ter and then add them togeth­er into a string. And because it was pulling these sort of ran­dom­ly from the file, you kind of resul­tant­ly got what looks like a ran­dom­ized string order.

And it’s a total mess, right? It’s a beau­ti­ful total mess. And I lat­er rewrote the code. It’s eleven lines long. It’s very effi­cient. But that’s the ver­sion that’s still run­ning online, because I just nev­er pushed the new one to the Linode. And it’s by far my most pop­u­lar bot, by many degrees, even though I’ve made things that are much more com­pli­cat­ed tech­ni­cal­ly after. 

And I think there’s some­thing kind of spe­cial in that, in that very very very occa­sion­al­ly, like once a year, some­thing goes wrong, and it will pull an unde­fined or a NULL or a NaN and push this to the Twitter feed. Which is not some­thing that would hap­pen if it was just a slight­ly more-together structure.

So yeah, as I’ve grown more pro­fi­cient I became real­ly thank­ful of this his­to­ry. It let me arrive at some kind of strange, unlike­ly places basi­cal­ly because I did­n’t know bet­ter or that my solu­tion was so bad that it com­plete­ly cir­cum­nav­i­gat­ed the ini­tial prob­lem and start­ed me on a new project.

As a cul­ture, we talk a lot about the naïve writer or painter or poet. But very rarely do we speak of the naïve pro­gram­mer. And I think that’s a real shame. I guess what I’m try­ing to say is that it comes through, you know. I’m an aura believ­er, in fair­ness, but I believe that the insides of a thing can nev­er real­ly be ful­ly invis­i­ble. That they intrin­si­cal­ly change how some­thing feels or behaves, even when they’re try­ing to hide. And would­n’t it be bet­ter to have the insides be inter­est­ing or unique, or non industry-specified, or built by some­one who’s kind of new and learn­ing the thing?

So, my last few min­utes I want­ed to switch tracks a lit­tle bit and talk about Twitter bots right now, and get more on a soap­box. And I want­ed to tell you you don’t have to make Twitter bots. And I’m sure you’re like, Duh I don’t have to make Twitter bots. I don’t have to make any­thing.” But real­ly, right? They’re rec­og­nized as a thing that cre­ative coders” do. People get press about them occa­sion­al­ly, like even big art blogs, Wired, or what­ev­er. And they’re won­der­ful in a lot of ways. They’re com­pact and man­age­able, and they can scale up to huge pro­por­tions or work small. There’s tools that help new coders get start­ed. And there’s vast APIs that are impen­e­tra­ble enough to even the most ardent developer.

But it’s pret­ty dang sat­u­rat­ed out there. And of course it’s nev­er impos­si­ble to come up with some­thing new and wild and rev­o­lu­tion­ary. And most of the peo­ple who are watch­ing this right now prob­a­bly have, prob­a­bly real­ly recent­ly. But it can be hard right now for some­one else to have not done it already, and I think I see these real­ly won­der­ful qui­et things get lost in the fog.

And you know, no big deal, right? One [?] can always do an idea sec­ond, or way bet­ter, and we’re all learn­ing, you know. And we’re learn­ing about how fab­u­lous it feels to knock some­thing out in a few weeks or days or hours, and how brevi­ty can great room for poet­ics to rush in. And watch­ing peo­ple inter­act with bots in such spaces is real­ly unique. They’re like a home for those strange mess­es of code, and for learning.

But still, let’s be hon­est. Branded Microsoft dis­as­ters might not be the end of the end. But it’s not the Wild West any­more. At very best it might be like a well-to-do pio­neer town. And at worst, maybe it’s on its way to like a sub­urb. But you know, this is all to say our skill sets scale. We know a bit of code, sure. But we’ve also learned to exe­cute ideas quick­ly. Prototype fun, tiny things. How to be sil­ly or seri­ous by turns. We’ve built com­mu­ni­ties and tools and made them viable and good for each other.

These skills do not just apply to cer­tain spaces and one com­mu­ni­ty on the Internet. All the things that have made Twitter bots this kind of ver­dant cre­ative space, the small com­mu­ni­ty of ded­i­cat­ed and pas­sion­ate weirdos, like the some­times loose terms of ser­vice, social plat­form that rewards small pub­lic inter­ac­tions, huge amounts of acces­si­ble human data… These are avail­able to any­one any­where on prac­ti­cal­ly any ser­vice or plat­form or app on- or pos­si­bly offline. So yeah, I guess I would ask you to take these tools and turn them to your own spaces. And yes, this is pos­si­bly Twitter, and prob­a­bly Twitter. But it prob­a­bly won’t be for­ev­er. And it might not be right now. And you don’t have to live there. 

So, I sin­cere­ly hope y’all, and I mean the cap­i­tal Y y’all as in the y’all that are mak­ing things seri­ous­ly right now for your present Internet, your com­mu­ni­ty in the world at large, build a fan­tas­tic, weird, tiny project renais­sance any­where that feels like it’s real­ly home.

So, thanks.

Further Reference

Video record­ing

Darius Kazemi’s home page for Bot Summit 2016.