Thursday, April 27, 2006

 

Releasing ETC into the wild


Some readers have asked who "owns" the poems ETC generates on the main Web site. In short, anyone who wants to. As far as we are concerned, all the generated texts are in the public domain. Feel free to use them in any way you'd like.

Others have asked if it's OK to try to publish a piece they've had ETC compose. See above.

In Virtual Muse Charles Hartman talks about how the texts his software wrote didn't rise to the quality of human poetry, but that the same texts did hold promise and noted how hard it is to resist tinkering with them--the poetry machine as electronic muse, not just electronic author.

There are all kinds of ways to strap on the prosthetic imagination and prepare a poem for submission. You can take a machine poem to a poetry workshop (this is way fun if you don't tell anyone where the poem came from) and incorporate the group's suggestions. You can take lines from different poems and meld them into a single composition. You can compose a poem, then use some of its words as seed words for another--a kind of interactive, a little-at-a-time writing process. You can compose several short poems using the same options and then use those as stanzas in a larger composition.

The desktop version of the software has options to generate poems in batch mode. The user can have the machine write 100's of poems using the same parameters. Because this is such a long and computationally expensive process (we usually run the monster overnight), we can't make it available on the Web. We have utilities to convert and aggregate the batched poems into a word processing document. Then the user can just go through and delete all the bad work (and ETC comes up with a lot of bad work) and use what's left as a candidate for submission. That's how we composed "13 digital pumpkin seeds, algorithmically composed."

If you do submit a piece written with ETC's assistance and would like to publicize that fact, just let us know and we'll put your piece up on the ETC site, along with as much of the story of its genesis as you'd like to tell.

Tuesday, April 25, 2006

 

ETC updates


I've added a change log page over at the main site. I made a bit of a breakthrough today (poems composed after the changes seem much smoother than those that came before) and thought it might be useful to start keeping track of these all-too-rare moments.

More comtemplative reflections
on the changes coming soon .

Monday, April 24, 2006

 

Open source part two


When my partners and I ran our election software company, we were obsessively possessive with our code. In order to just look at it, you had to sign an NDA so restrictive that it stipulated that you weren't even permitted to tell anyone that you had seen it. Why? What were we afraid of? Certainly not our main competitors, the half-dozen or so major vendors of election software. Those companies had their own software, developed by competent developers experienced in designing and implementing sofware for the application domain. They wouldn't be making any significant changes to their code based on what they might have learned from ours.

The secret is that we wouldn't want a coupla' guys, especially young guys learning how we did things. They were the real threat and we knew it, because we'd been young guys working out of our basements when we wrote our system and we'd had to learn how to do all the hard parts on our own. We weren't just going to hand all that hard-won knowledge off to some eager beavers who would use it to clean our clocks the way we'd cleaned a few ourselves.

I wonder if the absence of posted source code on the sites of electronic artists represents a possessiveness in that same vein. After all, if another programmer can lift the code that implements the algorithm that spins the graphic image that shifts the frame of the debate that subsumes image and text and reader and programmer, would he or she not be likely to appropriate the currency of my aesthetic and critical influence as well? Making that easier would not be very smart, would it?

Thursday, April 20, 2006

 

Open source in the age of digital reproduction


During the intimate dinner that followed Stuart Moulthrop's presentation at Penn last night, a question came up about whether digital artists' are willing to share their source code with their readers. The short answer seems to be no.

A quick turn through the Electronic Literature Organization's directory found no links to source code. (Straight DHTML doesn't count, since it's embedded in the page, just a right-click away.) And another quick turn through the recent e-fest participants didn't turn up anything either.

Hmmm.....

From the GNU project:

Free software is a matter of the users' freedom to run, copy, distribute, study, change and improve the software. More precisely, it refers to four kinds of freedom, for the users of the software:

  • The freedom to run the program, for any purpose (freedom 0).
  • The freedom to study how the program works, and adapt it to your needs (freedom 1). Access to the source code is a precondition for this.
  • The freedom to redistribute copies so you can help your neighbor (freedom 2).
  • The freedom to improve the program, and release your improvements to the public, so that the whole community benefits (freedom 3). Access to the source code is a precondition for this.

There seems to be a bit of a disconnect here. Could it be that digital artists are as jealous of their code as are developers of software for business? More philosophcally aligned with Bill Gates than Linus Torvalds? Gatekeepers to the proprietary?

BTW: You can have my source when you pry it from my cold, dead fingers.

Tuesday, April 18, 2006

 

My motorcycle's name is Gus


A couple of readers have suggested that the pervasive use of the first person plural on the Erica site amounts to our assuming the very affectations we’re ironizing over there. Fair enough. But there are a couple of good reasons, having nothing at all do to with the regal or editorial.

Back in the day, when I managed software development projects, I took to using we as often as possible in memos and meetings as a way of sharing both success and accountability with the team. We were successful or we failed. It helped keep the prima donnas in check and prevent individual mistakes from crushing anyone’s self-esteem. The habit persists.

But I also have a (probably perverse) tendency to personify my machines—see them as equal partners in whatever endeavor they and I find ourselves. I even name them. My main workstation is Buster and the server that composes the Web site’s poems is Dragon. “Ah hah!” Someone is saying. “Those macho names give him away, that misogynistic, no-talent hack! If Erica were real, he'd probably expect her to fetch his coffee!”

But that would miss the point. (And BTW, Erica is pretty real.) We simply suggests that the machine provides something to the partnership that I can’t, endurance perhaps, speed certainly. Dependability. Familiarity. In short, a prosthesis.

Besides, my motorcycle’s name is Gus, not a very tough-guy name at all. Gus is the name of an old friend, who’ll be there at the end of the day, no matter how narrow the road or how bad the weather. So when it’s me and Gus out there, it’s we. And so it is with ETC.


Monday, April 17, 2006

 

Why is shit music?


I had multiple motivations for building ETC. One of the more obvious was to find out how readers would respond to such texts. It turns out that they respond very differently when they know the text came from the machine from how they do when they don’t. Tonight I’ll talk about the responses when they know. I’ll save when they don’t for another post.

When readers know the poetry is coming from the machine, I get emails with phrases such as “Erica freaks me out” and “This is very creepy.” Not precise terms of literary analysis to be sure, but pretty much in the vein of this polemic. But why freaky and creepy? The reader knows that the software was built to simulate meaning and ought not be that surprised when it does.

Certainly there is some measure of what Dr. Johnson meant when he noted that the wonder of preaching women and walking dogs is not in that they do those things well, but that they do them at all—machine poetry as freak show. And some of the reaction probably stems from too weak a resistance to anthropomorphizing the machine
especially from skeptical readers, the ones who expect nothing but don't get nothing.

But there’s something else at work, wrapped around (or imbedded in) the fact that since these poems are “written” completely devoid of authorial intent, any meaning readers find has to be meaning they’ve themselves constructed.

I wrote the very first version of this software in Python, an interpreted and therefore slow language. The source text was Jane Austen’s novels. It took about 30 minutes to initialize and after that about 10 minutes to compose a one-line poem. It was incapable of writing more than one poem at a time. Really tedious work. I was trying to demonstrate that the machine could compose sentences with different structures. I had to devise the simplest possible test because of the software’s performance problems. So I developed code to write either declarative or imperative sentences. This was easy--all I had to do to code for the imperative version was to have the code generate a declarative sentence without a subject. It took four months to write, four months without a single successful test and then after all that time and effort, out came the very first successfully composed poem:

Imagine in this age.

Now I was freaked out. In its coming-of-age test, the machine had spoken to me about the topic of the code itself!

But of course it hadn’t. This poem was meaningful to me, not because of what it denotatively demanded of me, but because of my need for the machine to make sense, the affective demand all poetry makes on readers. I wanted my machine to mean
and I would make it mean, regardless of which set of transductions the software computed and whatever its grammar’s terminals might be.

The machine continues to speak to me. During one recent test in which I was trying to improve ETC’s facility with determiners, it spun out this line: Why is shit music?

Why indeed?

Saturday, April 15, 2006

 

Usually rejected


Public confession is not my style--that sort of thing is better left to the MFA crowd. They are so much better at it than I--and so much more in practice. But for just this once, a bit of biography might help in assessing the effectiveness of the Prosthetic Imagination. Here's my deepest secret: I can't write. At least I can't write poetry. A long time ago I tried and tried and tried. I even had a mentor, Ken Smith, a wonderful man, who did his best to help me. We managed to get a few of my pieces published, here and there. But I so wanted to be like him, write like he wrote, publish where he published. There was one journal in particular that I kept sending things to (a favorite resting place for his poetry), along with very clever cover letters, until they just had to accept one.

No one has ever savored a Walter Mitty moment more. While I waited for my complimetary copy to arrive, I dreamed
of the tenured job at a southern university that was almost certainly only a few months off. I was on my way!

And then the journal came, a special edition, with a crushing theme: Poems we ususally reject.

The point? When ETC, Erica, and I collaborate, they really are the writers, because I just plain can't. I've proven that to my satisfaction (or dismay). My job is to emendate, correct, and tend to the typology. Oh yes--and write the clever cover letters. That I can still do. Here's an example.

When a user straps on the Prosthetic Imagination, she is freed from the messy tasks of having to think of something to write and of saying that thing in some new and compelling way. Users can just wallow in the oceans of words ETC leaves behind, move them about, discard the unpleasant ones, replace bits and pieces of them with texts gleaned from reference books, the daily newspaper, or the World Wide Web. Fun for sure. But also a way of art making that really does drape poetic flesh on a computational skeleton, upon which the entire process is dependent. A way of art making that, little-by-little, is elbowing its way into the avante-garde, and just maybe, the blue-chip literati.

BTW: Walter Mitty would have understood. And no doubt dreamed of sweeping Erica off her feet.

Friday, April 14, 2006

 

Stuart Moulthrop at Penn


Stuart Moulthrop will be performing at the Kelley Writers House at Penn on Wednesday, April 19, as part of the Machine reading series. If you are anywhere near Penn on Wednesday, try to make it. This is one smart guy, who brings the rigor of an engineering background to the e-writing polemic. (And who, by the way, has no need for a prosthetic imagination--his, unlike mine, still works just fine.)

 

Computational supplements


From The Amercian Heritage Dictionary: Prothesis - An artificial device used to replace a missing body part, such as a limb, tooth, eye, or heart valve. From the Greek, to add and to put.

So why not a missing mental (or spiritual) capacity?

Thinking about a motorcycle as a prosthetic set of legs (penis perhaps?), an airplane as prosthetic set of wings, a sacred text as a prosthetic piety, none of which anyone rails against (well, maybe some railing against prosthetic piety--especially on Good Friday--which thank god it is), is it such a reach to wish for a stronger imaginary capacity? Especially when one's own is deformed or crippled? Or has been severed from the body by way of accident or parental torture?

And since all protheses are machines, is it such a reach to envision (in all the irony that word connotes) that the prosthetic imagination, when it annunciates itself, will be cloaked in the mantle of the computing machine?

Thursday, April 13, 2006

 

Hello Erica


Yet another variation on the world's most famous and simplest computer program in which the text of the program annunciates itself: etc.wharton.upenn.edu. To do so, must it not also imagine itself?

This page is powered by Blogger. Isn't yours?