The New York Review of Bots @nybots - Tumblr Blog

@TwoHeadlines: Comedy, Tragedy, Chicago Bears

BY DARIUS KAZEMI

I created @TwoHeadlines because of an email Adam Parrish sent me. Here is the thread reproduced in its entirety, with his permission:

Adam Parrish - Aug 23, 12:08 pm

Did you hear the news? Ben Affleck is the new CEO of Microsoft!

— Ian Bogost (@ibogost) August 23, 2013

^^^^ surely there must be a way to automate this kind of joke. the solution probably can't be purely lexical—I think you'd need to have some kind of semantic information about the people involved and the event that just happened. but it can't be THAT hard, right?

Darius Kazemi - Aug 23, 12:21 pm

I mean, I imagine you could just take the top 5 newspaper headlines and join them together by some kind of process, maybe markov-based or n-grams or something.

Darius Kazemi - Aug 23, 12:23 pm

If you go to news.google.com, you see this sidebar:

Top Stories NASDAQ Batman Steve Ballmer Damascus Mumbai Bo Xilai The World's End iPhone 5 Nidal Malik Hasan Justin Timberlake

Then click on one of them, like "Batman" and get the top headline:

The Guardian Ben Affleck to play Batman in next Superman movie

Then replace "Batman" with one of the other words in the sidebar:

The Guardian Ben Affleck to play iPhone 5 in next Superman movie

Adam Parrish - Aug 23, 12:26 pm

hmmm. yeah, that's not bad. arguably, though, you're generating weird headlines and not, like, jokes about current events.

At the time I thought Adam was wrong: of course @TwoHeadlines is telling jokes about current events! The corpus that it’s pulling from (news headlines) is nothing but current events!

Two months later and @TwoHeadlines is my most popular Twitter bot. I think it’s my funniest to date. Some representative tweets:

Wes Anderson's human rights violations continue with increase in public executions

Pentagon proposes USD 10.8 billion arms deal with NBC, UAE

'Elon Musk' returns, with more blood, revenge and a feisty makeover

Those cool United States Congress features? Android does that, too

Chipotle Mexican Grill is facing a new Islamist insurgency

Facebook, world powers report progress in nuclear talks, agree to further meetings

Then a few days ago, I saw this tweet:

Every funny tweet from @TwoHeadlines has been a magical realist Late Capitalist dystopia full of private armies and corporate states.

— Bones Demon-eyes (@BMunise) October 19, 2013

I did not set out to write a bot that writes near-future late-capitalist dystopian microfiction. I set out to write a bot that automates a particular kind of joke. But if we pause to consider the bot’s algorithm, it’s obvious where this tendency toward a very specific fiction genre originates.

The Google News sidebar described in the email thread above is Google’s attempt to parse out the subject of a bunch of related news headlines. For example, if there is a bombing in Iraq and there are a lot of news headlines about it, it will probably generate the subject “Iraq.” This is a very specific choice: it could have equally chosen “bombing” or “terrorism” or “chaos”, but Google’s algorithm tends to favor named entities over abstract concepts. What this means is that the subject of the news, as Google sees it, is almost always a corporation, a sports team, a celebrity, a nation, or a brand.

My algorithm builds its jokes by harvesting these subjects that Google has picked, and swapping them indiscriminately between headlines.

What is near-future late-capitalist dystopian fiction but a world where there is no discernible difference between corporations, nations, sports teams, brands, and celebrities?

So Adam was partly right in our original email thread. @TwoHeadlines is not generating jokes about current events. It is generating jokes about the future: a very specific future dictated by what a Google algorithm believes is important about humans and our affairs.

Who Led the Horse to Ebooks?

BY SARAH BRIN

You’re probably so sick of hearing about Horse_ebooks. You’re probably so over the whole thing, it’ll be a loooong time before you find yourself following another conceptual artist or majestic equine on Twitter.

Certainly, Jacob Bakkila's self-unmasking attracted much eye-rolling and lamentation from all corners of the internet. But you might be interested to know that his work in fact draws from an unexpected number of art historical precedents. Not just that lady who recently raised $600k to stare at other rich people all day, precedents worth knowing about.

I was able to reach Jacob Bakkila via email to discuss his work and learn more about his broader artistic practice. He has this to say about his influences:

Jackson Mac Low was my favorite poet growing up. I remember seeing a Christopher Wool painting when I was a kid called "CATS IN BAG BAGS IN RIVER" that really influenced my thoughts about language and symbols. Cory Arcangel is great, Ryder Ripps, a lot of people out of Eyebeam in NYC, Jayson Musson, really almost anyone who is making weird stuff online inspires me. Some webcomic people make really great stuff, Kate Beaton, KC Green, Brad Neely and a few others.

It makes sense that Bakkila would reference Jackson Mac Low, a poet associated with Fluxus; a loose affiliation of artists interested in playful multimedia art that eschewed the traditional gallery model. The same goes for his interest in Christopher Wool and his blocky text-based paintings, which often read like a stream of consciousness or Objectivist poetry. Does this look familiar?

Additionally, Bakkila references a range of artists working with new media, particularly Jayson Musson, the artist perhaps most known for his hip hop-inspired YouTube persona, Hennesey Youngman. In the series Art Thoughtz Youngman humorously (and astutely) critiques a range of topics in contemporary art history, including relational aesthetics, institutional critique and performance art.

In short, Bakkila’s influences check out. He’s citing artists working with play, with text, with humor, with performance, and with the internet. Bakkila situates his work at the intersection of several art historical traditions that imply the artistic intentions behind Pronunciation Book, Horse E_books _and _Bear Stearns Bravo are genuine.

However, some responses to the September 24th revelation articulate a sense of betrayal linked to the Bakkila’s day job as a creative director at Buzzfeed. But Bakkila’s occupation puts him in good company. Many (now) established artists have supported themselves by working in commercial media before their art careers took off--Andy Warhol, James Rosenquist and Ed Ruscha are just a few.

For others, Horse_ebooks and Pronunciation Book were tainted by their culmination as marketing for something called Bear Stearns Bravo. Even after spending a good amount of time on the project’s website, I had to ask Bakkila...what is _Bear Stearns Bravo?

Bear Stearns Bravo is a piece of video art with some interactive elements, heavily influenced by the CD-ROM games of the late '90s, Myst, MegaRace ... the piece's original codename was Night Trap. It's incorrect, but not entirely inaccurate, to call it a game, but there's not really a compelling definition of the word "game" these days, so I don't mind if people call it that: it's more like a choose-your-own-adventure-style art film. The story is a fictional retelling of the financial collapse of 2007/2008, mythologizing Regulators and Bankers as larger-than-life opposing forces, and it is within the context of this fiction that Horse_ebooks exists as a character, of sorts. While Horse_ebooks's tweets were drawn from the corpus of low-quality information products (ebooks) online, there's definitely a narrative across the account, and it was telling a story (in the way that a spambot would tell a story) that sets the stage for Bear Stearns Bravo.

(Still from Bear Stearns Bravo)

While Bakkila avoids using the term “game” to refer to Bear Stearns Bravo, the piece is surfacing at a moment in which non-traditional games have been gaining traction in the contemporary art world. Considering his prior work on the darkly humorous alternative reality game This is My Milwaukee, and the implicit humor seen across the spectrum of his art collective Synydyne’s work, it’s clear that Bakkila is no stranger to play. I asked him about what he finds interesting about working with games, or game-like material:

They're a fascinating medium, wonderful and terrible in equal parts, and almost anyone 30 and under has grown up with them as an equal to television in terms of things that vie for domination of the imagination, for better or worse. I read way more than I ever played video games, but the video games I did play, I played with consuming abandon. In my favorite games, a Fallout 2 or a Project Nomad, for example, I never really cared about "winning," or anything skill-based, I just wanted to look around and explore and experience the fictional world. Tom (Bender) totally agreed with me and you can see that design aesthetic in Bear Stearns Bravo: "wrong" choices are just as interesting as "right" choices.

It should come as no surprise that Bakkila is well-acquainted with the world of independent games. I asked if there were any he especially enjoyed:

I do not play a lot of video games these days, entirely because of time constraints. My favorite current game is ZiGGURAT on the iPhone, it's just about perfect. I play indie games & visit freeindiegam.es a few times a week. The designers thecatamites and kat_chastain both make incredible stuff. Porpentine's Howling Dogs is equally amazing, Emily Short does great interactive fiction, I love Ridiculous Fishing also on iPhone, Anna Anthropy's game Redder. I'll probably buy a DS and play the new Animal Crossing now that I have some more free time.

While the mystery of Pronunciation Book has been solved, and the non-robot brain behind Horse_ebooks has been exposed, there are reasons to remain excited about what’s to come. Even if you were totally nonplussed by Bravospam or Bear Stearns Bravo, it’s worth noting that we’ve reached a new milestone for public art. While artists have been working with Twitter and YouTube almost since the platforms’ inceptions, the Synydyne projects are arguably the broadest-reaching artworks executed through those media. The public nature of those channels has subsequently created a very public discourse--a multinodal discussion, sometimes charged by outrage, interest or a self-establishing fan culture. I am hopeful that conversations like these will continue to happen, everything else seems to.

The End of Horsebooks Is Hardly the End of Anything

Let's talk about what happened today! A bot, long suspected not to be a bot, ceased bot-operations and dramatically unbotted itself in one of humanity's most genteel, least mechanical forums. Concurrently, its creators appeared in a lower-Manhattan gallery to {sound of the channel changing}.

Look, their message was clear: ART. They did an art, and would seemingly like to keep doing arts. They arted so hard that Susan Orlean is writing a story about them for the physical New Yorker, a feat of digital-to-analog enshrinement most bots only dream of, to the extent they have been programmed to dream.

(If I understood tumblr footnotes better I'd go on a digression here about how Susan Orlean, who has seen herself portrayed in one of our more enjoyable self-referencing films, is just the right tautological lens for the more tail-consuming aspects of this story. Better yet, future-pretend I'll write a whole thing about it the day her article comes out.)

Understandably, today's unmasking/unplugging vexed a lot of people who'd imprinted on the shared hallucination of this machine gone adorably haywire. Some (and this admittedly falls between subtweet and strawman) fixated on where its puppetmasters keep day jobs. But that's a pretty facile dismissal of what looks to have been a sophisticated operation, and chucking this as churn-inflected marketing only boosts the odds you'll fall for their next thing, too.

On reflection, I can't sustainably hate anyone who spent two years scrimshawing cred from the loose fingerbones of Internet Spam Voice. This morning put me in a kind of snarky bafflement, tinged by admiration that someone committed to a bit this ridiculous for so long. I liked Dan Sinker's gobsmacked elegy -- not because it was someone else who ended their popular thing getting a taste of their own medicine, but because it marked another notch on an increasingly crowded wall of cultural experiments bending the super-weird line between person and process.

So let's not pretend the revelation of this antibot means the Internet has stopped being history's greatest serendipity engine. Or that bots, the little scripted dingbats, have lost their capacity for channeling that engine to often miraculous effect in social contexts, to taste. It's probably true Horse was the most famous bot, but also probably true it wasn't nearly the best.

In losing @horse_ebooks, readers, we have lost a pioneer. But we are left with its America.

Welcome, Humans

Every time some enquiry began in earnest and then ditching 30 minutes later. #badass #swag

— Latour Swag (@LatourSwag) April 5, 2013

The New York Review of Bots is proud to announce the appointment of two of an unspecified number of new Senior Editors, Tully Hansen and Ben Abraham. They will jointly oversee our Australian bureau.

It goes without saying that this is a major win for all NYRB stakeholders, be they human, scripted, or elsewhere on the algorithmic spectrum.

Testing the bot-program

Oh hello

BREAKING: @PowerVocabTweet Launched

Adam Parrish has released a new bot, @PowerVocabTweet, which he describes as "@everyword's dada cousin." Every two hours, it posts a randomly generated word and invents a definition for that word. English is the better for it.

flerticulative, adj. constituting or like a gerund

— Power Vocab Tweet (@PowerVocabTweet)

April 18, 2013

Ironically, the word "flerticulative" is not itself flerticulative.

scruve, adj. of or relating to an uncle

— Power Vocab Tweet (@PowerVocabTweet)

April 17, 2013

Clearly scruve has the inside track on avuncular, especially for haiku artists.

Awer-Normandy, prop. n. medium-sized wildcat of Central America and China

— Power Vocab Tweet (@PowerVocabTweet)

April 15, 2013

Ladies and gentleman, your codename for OSX 10.9.

(follow @PowerVocabTweet)

Why do we let bots be so violent?

— Best Frienjamin ❤ (@10rdBen) April 9, 2013

new do and so the prophet became king.

— Latour Swag (@LatourSwag) April 13, 2013

a partial list of nonbot entities whose bot-counterpart could be eponymous: Meg Cabot. Radbot, Count of Hapsburg. Ribot (horse). Tanzverbot.

— Adam Parrish (@aparrish) April 6, 2013

6899 Nancychabot. Łukasz Kubot. US senator Isham Talbot. Rabot Island. Bot, Tarragona, Spain. (source: wikipedia en.wikipedia.org/w/index.php?ti…)

— Adam Parrish (@aparrish) April 6, 2013

timeshaiku

A haiku from the article: Isabel Allende: By the Book

Pasteur did not wait for his ideas to emerge fully formed from his laboratory and get back into bed. #swag

— Latour Swag (@LatourSwag) April 6, 2013

We pull from the homepage — but no opinion articles, no sensitive things. I don’t want to have it pull something on, like, North Korea, because I don’t want people to think we’re making light of something serious. So from there there’s about 75 articles to work with — and from there there’s about 15 potential sentences to work with the syllable count on a given day.

http://www.esquire.com/blogs/culture/new-york-times-haiku-generator#ixzz2Po91JQuh

Sunday reading from the week that was: Esquire interviews Jacob Harris, brainparent of the NYRB-featured New York Times Haiku Generator.

The nytimes.com top-level domain is for bots what the Times' op-ed page represents to many humans: a place of prestige, a trapping of legitimacy. And yet: perhaps this visible perch, and the expectation of respectability that comes with it, subtly erode the bot's poetics. After all, by their very controversy, one would expect "opinion articles [and] sensitive things" to generate more striking Haiku.

So who's going to start the too-hot-for-nytimes.com NYT Haiku site?

Meet @iederwoord: Dutch @everyword clone and aspiring standup

budgetbewaking

April 4, 2013

First time logging into Hotmail in ages. Noticed this epic spam headline: "We apologize, it wasn't Stryper's fault. Great Stryper Article.‏"

— Jason Killingsworth (@jasonkill) April 4, 2013

I don't tag and what is modern?

— Latour Swag (@LatourSwag) April 4, 2013

This is not a bot; we are the bots

I'm an Equal Opportunity Offender, by which I mean you all appear as one faceless subhuman mass through the lens of my Universal Perspective