Thank you and good day.

For all of you listeners out there who just can’t wait for ep1 to drop:  have patience!  It will arrive soon!

• Go to Craigslist
• Do a search
• Copy the URL
• Go to Earthify Beta
• Paste the URL
• GO!

All Hail Jesus’ Name:

All Things are Thine

Crown Royal

Jesus Loves Me

My Faith Looks Up to Thee

O Come, O Come, Emmanuel

Onward Christian Soldier

Prelude and Fugue in G Major – Fugue

Prelude and Fugue in G Majog – Prelude 

Paslm 19

Savior, Like a Shepard Lead Us 

St. Annes Prelude

Trio Sonata No. 1 in E major

Trop Sonata No. 5 in C major

Trio Sonata No. 6 in G Major

or Download the whole thing!

Crowded is an montage audio program similar to radio shows like This American Life, The Moth, or the productions of Joe Frank.  What makes it unique is that all of the material is is made up of segments of audio requested from and submitted by workers on Amazon’s Mechanical Turk site in return for payment.  I have made one incomplete episode as a test.  You can listen to it here:

Crowded – Episode 0 – (beta)
(I apologize for the audio quality of the first story – you have to listen very closely to hear what he is saying.  I will fix it eventually, but this was fine for this proof of concept)

Each episode has a mechanism or a theme.  For the test episode, I asked workers to listen to a collection of songs.  After listening, they picked one that reminded them of an important event in their life, and then recorded themselves telling that story.  You can see the HIT (Human Intelligence Task – the instructions given to the workers) at the bottom of this page.  The tentative list of future themes are all listed below in the prouction timeline.  Some of them have nothing to do with labor or MT.  Some are very self-conscious.

The workers are given around $5-$10 to follow the directions like the ones below, depending on the complexity of the task. The jobs usually involve recording some audio. Depending on the assignment, they will use either a custom-built recording applet or a commercial conference call service.  When I receive the audio, I (with the help of Graham Reznick, an audio engineer) organize and edit the results into a professionally produced radio show.  Although it would be interesting to keep these results as raw as possible and strive for objectivity, I have decided that I would rather focus on creating an intriguing show.  Except where the theme specifically calls for it, I will not give up editorial privileges.

Why?

After doing several projects using crowdsourcing (Dirt Party, Invisible Threads, You’re So VIP) that essentially use workers as a faceless labor force only slightly better than a computer, I wanted to do a project that was about the actual people who are doing these tasks.  Who are they?  Where do they come from?  Why do they do these jobs?  The dynamics of “buying” a story via Mechanical Turk are interesting to me.  Do I own the story?  What kind of stories do you get from workers  who are ultimately looking at the estimated time vs. payment bottom line?   You could argue that I am still not humanizing them – that I am simply exploiting their willingness to tell personal stories for a few bucks.  But this project isn’t really taking a position in terms of the ethics of online labor markets.  It’s simply the MT system – which most believe is a kind of foreshadowing of a labor paradigm that will inevitably become much more common in the future – in a way that I think is more interesting and rewarding than most of what I have seen.  Most of the workers who respond to my requests are very excited to have been asked to do something that requires a little more creativity and thought.  The trick will be to bring out something unique about the kind of people who do this kind of work.  The average person doesn’t know what crowdsourcing is, and especially doesn’t care to hear strangers talking about uninteresting topics.  So the concept alone won’t make this project successful.  The tentative themes below are intentionally being left somewhat vague so that, when I gain a better understanding of the workers , I will be able to craft a show scenario that will coax better material out of them.

The end result will be a CD of these shows along with a small pamphlet or booklet about the workers who contributed, and any thoughts or research that I accumulate over the course of the show.  Because I don’t want people who wouldn’t normally be on MT to accept the tasks, I will release the entire collection after the project is done.   I will also release the MP3s of the shows and all of the source audio on the web.  With the help of Rhizome, I hope to one day hope to pitch this to WNYC as a series that they can play on air.

Project Budget & Timeline

I would like to make 1 episode per month for 1 year, starting in September of 2010.

Episode Budget

• $100 for stories (roughly 5-10 workers @ $10/job to get enough usable content)
• $200 audio editing
• $50 audio/hosting services (conference call lines, media server, etc.)
• TOTAL: 1 episode costs around $350

Main Budget

• 12 Episodes: $4,200
• Jingle composition: $800
• Printing booklet: $300
• TOTAL: $5500

Timeline

I will complete one episode each month and send it out to a small list of people for comments.  The following are the tentative themes for each month.

1. September Listen to a series of songs, choose one that reminds you of an event in your life, and then tell that story.  The stories will be collected and played on top of the song that you choose. (11 calls @ $8)
2. October Answer the following questions (20 calls @ $4)
   1. describe your immediate environment
   2. how/when did you working on MT?
   3. how do you plan on using the money that you receive for this work?
   4. record 2 minutes of ambient audio from somewhere nearby
3. November Call me and have an argument about a topic I give you.  I will edit myself out, leaving only your side of the argument. (5 calls at $12)
4. December Call this conference line.  You will be joined by 2 other workers.  Your character is [fill in a character].  During the course of this call, you must [plot point 1], [plot point 2], [plot point 3].(I’m not quite sure how I will orchestrate this one yet, so it is very tentative – 6 calls @ $15)
5. January Use the recorder below to record yourself answering the following questions:
   1. Who do you think I (the Requester) am?
   2. Do you ever think about the people who give you the tasks that you do on MT?
6. February Use the recorder below to record yourself reading from your favorite book. (12 calls @ $5)
7. March Use the recorder below to record yourself talking about previous jobs that you have had. (10 calls @ $8)
8. April Use the recorder below to record your first name, your last initial, and where you live.  (This show will focus on quantity rather than quality.  250 calls @ $0.5)
9. May TBD
10. June TBD
11. July TBD
12. August The ubiquitous “best of” show.

Resume or Curriculum Vitae

Please see my CV here.

Work Samples

Invisible Threads

Dirt Party

Praying@Home

You3b

Sample MT HIT

Tell me a story

I am working on a radio show called This Mechanical Life where all of the material comes from Mechanical Turk workers.  And I need your help!  Just follow the 6 steps below.

• Listen to the songs below.
• Think of an event in your life that reminds you of that song.  This could be any story, but it should be important to you, or something that you think other people would enjoy hearing.
• Find a quiet room where you won’t be disturbed
• Tell me your story! Follow the instructions below to record the story. Some tips:
  • Please dont mention the song that you chose while telling your story, or even that the story you are telling was inspired by a song.  Just tick the circle below and the song will play underneath your story.
  • You can use either your computer microphone or a telephone.
  • The story should be between 8 and 15 mins.  Please try to stay on topic and don’t ramble.  That being said, just relax and don’t rush through it.
  • Speak up, speak slowly, and speak clearly.  No bonus will be given if I can’t understand what you are saying.
  • Do not play the song while you are recording.
• Tell me how you recorded it here.
  • if you called the phone number, put the number from which you called here.
  • If you used YouTube, put the URL here
  • If you used HoundBite, put the URL in here
• If you wish to be credited in the final production, please enter your information here.  I cannot guarantee that your story will be used, or that it will not be edited or re-told.  But don’t let this dissuade you!  If you have a problem with any of this, let me know and I will work with it.

The Songs (also available at http://4u.jeffcrouse.info/mt/)

	This text will be replaced	Freescha – Holiday Frost
	This text will be replaced	The Dead Texan – Girth Rides a (Horse)
	This text will be replaced	DNTEL – Last Songs
	This text will be replaced	Julian Fane – Exit New Year
	This text will be replaced	Microstoria – Edu
	This text will be replaced	The Octopus Project – What They Found
	This text will be replaced	Secede – Big Day Out
	This text will be replaced	Am-Boy – Turning Of Season
	This text will be replaced	Neoangin – Trippy Disco
	This text will be replaced	Ratatat – Brule
	This text will be replaced	Ratatat – Bird Priest

Recording Options

• Leave me a message. The maximum length is 3 minutes. You will have to call back a few times.
• Use YouTube QuickCapture.  You don’t have to even be in the frame as long as the audio is clear.  The maximum length is 10 minutes.
• HoundBite is like YouTube for audio.  So if you really don’t want to record a video, this is your solution. The maximum length is 15 minutes.

The trip went very well overall, and I hope to do it again soon.  The wonderful Christina Kral is currently working on a video about the trip, and here is a small preview.  Stay tuned for the full version.

draft from christina kral on Vimeo.

We had a great time in Iowa.  Everyone was super nice, and we built an awesome fort in the Studio Arts Building, which was, awesomely, a former Home Depot (but they call it Manards in Iowa — silly Iowans).

Steve Lambert and I did a “The Web for Artists” workshop.  Jamie O’Shea and Christina Kral did a circuit bending workshop, and Steve did a Shopdropping workshop.  It was a great start to the Roadshow.

I spent the weekend in the editing room making some new videos for an application.  See them after the cut.

The rest is history.

• Wired – September 19, 2008 | 9:09:55 AM
• Digg – September 19, 2008 | 11:00 AM
• BoingBoing – September 19, 2008 | 12:54 PM

I saved the BoingBoing and Wired for posterity.

paid-placement-for-brands-in-pop-lyrics-boingboing

products-placed_-how-companies-pay-artists-to-include-brands-in-lyrics-listening-post-from-wiredcom

Others

• Which Rappers And Divas Get Paid For Product Placement?
• Yes, Ludacris will sing about your jeans
• Selling out: Product Placement in Songs.
• The First Rule Of Product Placement In Songs: You Don’t Talk About Product Placement In Songs and here | September 19th, 2008 at 9:24 pm
• And The Brand Plays On Pt. 12
• Ad or Art? Why Not Both? | Sep 19 2008 2:32PM EDT
• Products Placed: How Companies Pay Artists to Include Brands in Lyrics
• Ad Supported Music: inside the ‘secretive market for song lyric product placement’
• Product Placement in Songs
• Anti smart | 5:59 PM on Fri Sep 19 2008

Polls have been taken. Phone calls have been made. Hands have been shaken.

The result is this:  The Eyebeam Roadshow will be traveling to the Chicago area in late October/early November for a short warmup tour.  Then, in mid-November, we will be heading out to California, starting off in San Francisco, and then heading down to LA.  We will be at the following venues:

(maaaybe) Columbia College
(probably) University of Illinois
(probably) Iowa State
(probably) The School of the Art Institute of Chicago
November 15th: Mills College
November 16th: ??
November 17th: Drive from SF to LA in our rock ‘n roll tour minivan
November 18th: UCLA
November 19th: UC Santa Barbara
November 20th: Cal Arts

Right now, the lineup is:

• Steve Lambert
• Andrea Polli
• David Jimison
• Michael Mandiberg
• Friedrich Kirschner
• Christina Krall

For the last few weeks, I have been working with Friedrich Kirschner, Artem Titoulenko, and Sophie Dutton on this projects. To be honest, I had very little to do with the actual coding of the project, which is the impressive part. But I did put together the Finger Olympics website WordPress Theme and make some videos with the Finger Olympics gang.

This time around, we decided to make a few key changes in our strategy, based on our experience at the Eyebeam benefit.

1. Simplify the output.
2. Simplify the system.
3. Have some backup plans if the tech fails.
4. Don’t have “roles” – that is, don’t separate the work of creating the “dirt”

So what we came up with was a system where we invited 3 groups of people to create LOLCat-style captioned photos of people at the party based on some very basic autobiographical information.

The first group was “the Insiders” – people who we had invited to use the futuresonic.dirtparty.org site and joined us in an AIM chat room. Dave and I would insert people into the system, and then tell the group to attack,

The second group was the people at Futuresonic. Once again, the trusty Eric Olson came through, creating some amazing kiosk software so that anyone at the party could walk up,

Source Code

As always, I haven’t really fully cleaned up and commented this code, but hopefully it will be of use to someone. If you do want to have your own Dirt Party, please let me know and we can work out how to make the software work for you.

The Backstory

Ever since I discovered internet radio back in the days of Live365, I have been somewhat obsessed with having a station of my own. My first station was Quadraphobia Radio (which is somehow still listed on the Live365 site). I even had a weekly live show with my friend Misha that I think a total of 5 people ever listened to. Quadraphobia Radio died with Live365 went commercial, and for a time, I was without a station. A2L2 was born when I rented my first private server from ServerBeach, and for a while, it thrived, getting up to 10 daily listeners, which was just fine by me. But after a few devistating hardware failures at ServerBeach, I moved over to Dreamhost, but I didn’t have a dedicated IP – that is, until Dreamhost PS. So now everything is right in the world, A2L2 is back up, and hopefully it will stay up.

I think I am going to make a new scheduling system and maybe even have another live-ish show, so if you want to be informed, or just talk about experimental/IDM music, join the new A2L2 discussion list.

Okay – I was rushed.

And the audio on this one is really off… We’re working on it.

Dirt Party Kiosk

This software was running on a kiosk machine at the Futuresonic Dirt Party. It basically allows people to come up and dig up dirt about people they know and create the LOLCat-style captioned photos that get sent to the central repository.
http://svn.digitalsituations.com/dirtpartykiosk/

Anywhere But Here Uploader

This is another kiosk application that allows people to insert a card reader with photos and send them out to Flickr, PhotoBucket, Facebook, and email.

http://svn.digitalsituations.com/photoblaster/

Steve Lambert now joins the BoozBot Operator Roster along with Scott V. from Oakland. Steve sent this chat transcript, which is a bit hard to read because it is only one side of the conversation, but funny nonetheless.

David Jimison, my collaborator on the project, also posted his thoughts about the project and the evening.

How He Works

My role in the creation of BoozBot was mostly on the software side, with Dave doing most of the building. The main questions that interested me were:

1. How can I make software that contains its own personality, even though the operator might change?
2. How can I make a text interface that is extremely easy to use, and minimizes the disadvantage of having to type in responses in a real-time conversation?
3. Where is the sweet spot between automation and human intelligence?
4. And, one of my favorite new questions: are there any general principles for writing funny software?

So I thought I would give a little explanation of how he currently works, and how I answered each of these questions.

First of all, for the sake of simplicity, we use puppet terminology to talk about BoozBot. For instance, the person who is controlling BoozBot is the puppeteer, BoozBot is the puppet, and the software that runs on the laptop attached to BoozBot is the puppet software. We decided pretty early on that we wanted to try to minimize the requirements for the puppeteer by not requiring any additional software beyond a simple Skype client, so all of the functionality for BoozBot exists on the puppet side. This made for some interesting interface challenges.

The technical requirements for the puppet software were that it be able to

1. receive Skype video calls from a puppeteer
2. receive Skype messages from that puppeteer and interpret them
3. read text aloud using a text-to-speech engine, and fire viseme events to trigger mouth shapes
4. send serial commands to an Arduino board in response to commands from the puppeteer
5. take photographs from “eyes” and send them to Flickr

After much research and deliberation, I reluctantly decided to go with C# for this first version of the software. At first, I was determined to use an open language. I wanted to develop it in OpenFrameworks as a mac-only application. I wanted to figure out how to get access to the Carbon Speech API, and use the Skype.framework with the wonderful iVox voice fonts, but it ended up being way too difficult. I couldn’t figure out how to create wrappers for the Objective C code for C++, and I didn’t have the time to really dive in and figure it out. C# just makes stuff so damned easy, and I had already bought a bunch of AT&T NatrualVoices voice fonts, so I eventually caved in. Furthermore, in the end, we decided that Microsoft Mike – one of the default Microsoft voice fonts – was perfect for our needs since BoozBot was supposed to be kind of low tech.

For the 1.0 version, I plan on re-writing the software in openFrameworks or Java, depending on a little additional research.

The full list of resources I used:

1. Visual C# 2008 Express Edition
2. Skype4COM
3. EmguCV (although the features aren’t yet active)
4. DirectShowNET a C# wrapper for DirectShow
5. Microsoft Speech API
6. Flickr.NET with help from the Coding4Fun tutorial
7. ForecastXML service from Weather Underground AP

The following list is the commands that the puppeteer could send to boozbot. To connect to BoozBot, the puppeteer would sign onto Skype and simply call the ‘boozbot’ user. BoozBot would automatically answer the call (if you are on his friends list) and start sending the puppeteer a video feed. The puppeteer can then send boozbot any text or command he wishes. The commands work just like IRC – anything that starts with a ‘/’ is considered a command, and everything else is dialog.

1. Any text sent to BoozBot that does not begin with a slash (‘/’) will be treated as dialog and read aloud.
2. /list, /l Prints out a list of available commands
3. /p This puts BoozBot into “processing mode” (a whirring, ticking sound starts and BoozBot says “transmission recieved. processing”, and should be sent immediately after a customer finishes talking and expects an answer. This will buy you time to type in a response.
   
4. /pour [drink#] Without the drink#, this command sends you a list of the available drinks and then waits for you to pick one. If you do provide the drink#, it pours the specified drink immediately.
   
5. /shutup, /s Stops whatever BoozBot is saying.
6. /repeat Repeats the last thing BoozBot said.
   
7. /blink, /b [optional blink#] Without the blink#, this command sends you a list of the blinks that BoozBot can do and then waits for you to pick one. If you do provide the blink#, it executes the blink imediately.*
8. /smile Makes BoozBot smile.*
9. /frown Makes BoozBot frown.*
10. /voice [optional voice#] Without the voice#, this command sends you a list of the available voices and then waits for you to pick one. If you do provide the voice#, it switches to the voice immediately.
11. /macro, /m [macro#] Without the macro#, this command sends you a list of the available macros and then waits for you to pick one. If you do provide the macro#, it performs the macro immediately.*
12. /time – Makes BoozBot to read out the time.*
13. /pic - Grabs a still from the video stream and sends it to its Flickr account.*
14. /weather - Makes Boozbot read out the weather forecast for New York.*
15. /turn [degrees] Tell BoozBot to turn his head.�
16. /sports – reads some sports scores for the local team.�
17. /quote – Makes BoozBot read out a random wise quotation.�
18. /parrot – repeats the audio that has been recorded since BoozBot last finished speaking.�
19. /remember [name] Remembers someones name for use in later automatic speech.�

* These commands will be called automatically if no command is sent to BoozBot for more than 45 seconds.
� Not yet implemented. Included in this document to build excitement.

These features are also covered in the BoozBot (beta) handbook. The macros were a big breakthrough in terms of developing a character using the software. The repetition of phrases made BoozBot seem more robot-like, which I think made people less sure that there was a real person controlling him.

Programming the puppet software was just a matter of wiring up the Skype API, the Microsoft Speech API, some webcam stuff, and learning a very little bit about serial ports, and then just writing a small command interpreter. The hardest part by far was figuring out the DirectShow webcam stuff. I ended up using a program called WebCam Splitter by Very Soft to split the signal coming in from the webcam so that I could grab frames while Skype still had access to the video feed so that the puppeteer could see his customers.

For some reason, I never could get OpenCV to recognize the “virtual webcam” created by WebCam Splitter, which is why I had to resort to using DirectShow to get a list of the video capture devices. The examples on the DirectShow .NET Sourceforge page show you how to do just about anything you want to do, with the exception of gaining access to the individual frames coming from a webcam. As a result, I never got EmguCV/OpenCV playing nicely with my webcam.

Source Code

Here is the Subversion repository for the puppet software. It comes with all of the DLLs, so I *think* the only thing you’d have to install is Visual C# Express.

http://svn.digitalsituations.com/boozbot/

Future Plans

1. “most commonly used commands” command, which will query a remote database for the most frequently used dialog or commands by any operator
2. “recent commands”, so that operators have access to their last X commands for easy retrieval
3. more complicated macros that include movement/blinking/etc, and text replacement macros, so that the operators can plug words into existing macros.
4. add the ABSML parser as an additional web service so that this puppet software can be used for James Chimpton, just like the weather or news commands
5. Eventually, we’d like to have a family of BoozBot-like puppets in different bars. Although this will be some time in the future, we will be preparing for this by building features that will react to a face entering the frame by pinging any registered puppeteers who is logged onto Skype and invite them to control BoozBot. We will also build in some kind of payment system, so that the puppeteers will be paid for their time.
6. Make BoozBot more self-contained – smaller hardware incorporated into his body. Soon we will be sending out a request for an intern to help us with building the new body.
7. Fake beathalyzer
8. more daemon/automatic activities
9. fake face/ID scanner

A few months ago, Steve and I were informed that Turbulence awarded us a grant to finish ABSML – the markup language that started as Fartsy – the Artist Statement Generator, then evolved to ASSML – Artist Statement (extra S for Savings) Markup Language, then had a short stint as DPML, or Dynamic Prose Markup Language, and is now ABSML (pronounced “abysmal”), which can stand for a number of things, but probably most officially, “A BullShit Markup Langauge”, which is fitting on many levels. I have written about it previously here and briefly here.

This happened after ABSML was used to power James Chimpton, and had inquiries from a few other organizations to use it as a kind of spam-blocker-buster for political activism. I have mixed feelings about bringing something into the world that could also be used by spammers with less honorable intentions, but as the language is evolving, I’m starting to see that it is moving away from something that could be used in such a way.

After talking with Angela Ferraiolo today, it’s clear to me that this will be an application that requires more interaction. It shouldn’t be so much click and go, automatic text generator. Rather, it should involve some human decision making at the time of parsing. This makes it much more interesting for performance, and I’m sure the results will be richer. So kudos to Angela for giving us a new way to think about ABSML and breaking us out of the rut that we had been in for about a year now.

Aspiring artists take note – after the break is a grant application was written by a master of the form – the incomparable Steve Lambert.

Steve Lambert
slambert@eyebeam.org
602 Grand St. #3
Brooklyn, NY 11211

http://visitsteve.com

http://whytheyhate.us

http://antiadvertisingagency.com

http://www.addart.eyebeam.org

Steve Lambert was born to a former Franciscan Monk and a Dominican Nun who practiced a variation of Liberation Theology, then abandoned organized religion and left the church a year before his birth. Despite never graduating from high school, Steve went on to study sociology, film, and music before receiving a Bachelor of Fine Arts degree from the San Francisco Art Institute in 2000 and a Master of Fine Arts degree at University of California in 2006. After years of doing street-level interventions, he founded the Anti-Advertising Agency in 2004. In the past, Steve has worked as a furniture installer, radio host, record store clerk, ballet dancer, parking lot attendant, undercover store investigator, theatre house manager, delivery truck driver, upright bass player in country western band, school teacher, landscaper, and lecturer among other things. He currently claims artist and professor on his taxes.

Steve’s projects and art works have shown throughout the United States, as well as Cuba, Canada, Holland, and Barcelona. Writings about his work have appeared in multiple publications such as the New York Times, Punk Planet, El Pas, and Newsweek Magazine.

Steve is currently a senior fellow at Eyebeam’s Open R&D Lab in New York where he is developing software that will block ads on websites and replace them with art.

Jeff Crouse
jeff@eyebeam.org
47 Maujer Street apt 2D
Brooklyn? NY? 11211

http://jeffcrouse.info

http://www.doublehappinessjeans.com/

http://you3b.com

http://www.realtimeart.com/switchboard/

Jeff Crouse is a digital artist and programmer currently based in Brooklyn, NY. He creates art that uses live data sources. Jeff received has a bachelor’s degree from NYU’s Gallatin School and a Masters of Science from Georgia Institute of Technology in Information Design and Technology. His work has been shown at the Sundance Film Festival, the Come Out and Play Festival in Amsterdam.

Collaborating with Stephanie Rothenberg, Jeff created the first sweatshop to be built in Second Life. Another project, Earthify, is an online application that maps a set of geographical data onto Google Earth for easier Earth-based browsing. Jeff is the developer of Switchboard, a Java library for Processing that allows artists and designers to easily use a variety of web and network data sources such as Flickr, Yahoo!, Google, and Amazon in their work.

Jeff is currently a Senior Fellow at Eyebeam Center for Art and Technology and is working on a WordPress plugin that will collectively prevent anything from ever being deleted from the web.

PROJECT PROPOSAL
DYNAMIC PROSE MARKUP LANGUAGE

Dynamic Prose Markup Language (or DPML) is a new markup language that we have been developing for the past year in the Eyebeam OpenLab. Surely you are already familiar with HTML or Hyper Text Markup Language, and XML or eXtensible Markup Language. These languages allow us to write code that make up the web as we know it today. But like it or not, the internet is boring and dumb.

Dynamic Prose Markup Language is different because it’s a language which writes itself, thereby making the internet interesting and smart. DPML could be considered a text generator, but it’s much more sophisticated than that. How does it work you ask?

DPML inserts text into sentence templates. As we learned in elementary English, all sentences can be broken into component parts. For example:

<Greeting> <noun>, are you ready to <verb>?

could become:

Good evening, Iowa City are you ready to rock?

or

Hey, McEnroe are you ready to serve?

DPML includes a variety of tags in place of these parts of speech and sentence components. In the right combination, the tags create prose which, while based on formulas and code, do not appear formulaic.

TAGS
To introduce variety, DPML uses the internet to compose it’s prose. For example, if one were to use the <synonym> or <antonym> tag, DPML uses online dictionaries to find synonyms and antonyms, in the correct part of speech, and insert them into the document. Perhaps you want to name someone topical in your generated document? With the <person> tag DPML will scour a website for name, choose one at random, and include it in your document.

For example

<person url=”http://www.nytimes.com/national/”>

will rip a name from the national headlines and insert it into your document.

This is a sample of a few tags and sentences they fit into. DPML can also be programmed to order these sentences into paragraphs, and randomly order the paragraphs, all before the document is delivered. These steps insure that the formula used to generate the text will be indiscernable to the reader.

WHY?

Yes, at this point you might be wondering; what’s the point of creating a new markup language to generate text on the fly?

At first it was just a fun project that got out of hand. Having read our share of obtuse artist statements we wanted to make an Artist Statement Generator that would blab on for a couple paragraphs about “notions of the material/non-material dialectic” or “facilitating the energized application of theoretical understandings.” We were pretty sure we could do this with a computer. And almost as sure that if we did, it would be funny. Basically we were making a bullshit generator for our own laughs. We wanted it to be more than that, but it just wasn’t.

Then Sam Gould from Red76 came by and asked if he could use DPML for his “Befriend a Recruiter” project. He realized DPML could quickly create letters of interest for military recruiters from bogus potential enlistees, thus keeping the recruiters busy writing follow up emails to people who had no interest in joining the military.

Later, Neighborhood Public Radio visited and asked if DPML could be used in their pirate radio project, “American Life” at the 2008 Whitney Biennial. At the time of this writing we are adapting the software for a radio program called “Inside the Artist’s Studio” hosted by a robotic Chimpanzee named James Chimpton. Chimpton’s brain is DPML programming sent to a text-to-speech generator which voices the robot. The result is rather absurd — a clothed robotic monkey, interviewing artists in the Biennial about their work, in a robotic english accent.

WHAT THE TURBULENCE COMMISSION WILL PAY FOR

After a year of development, the Turbulence commission can help us release DPML for public use and open source development. We would use the $4000 in funds in multiple ways. The funding would allow us to hire programming assistants to develop an API so that DPML can be used in other applications. The Turbulence Commission could also allow us to offset the cost of building a front-end so the application can be used via the web. And funding could allow time for us to better research how others might want to use DPML and change it accordingly. As an open-source project , we would open the DPML standard up to the public by holding calls for recommendations (like a shadow W3C). These recommendations would then be implemented into a standardized parser.

While beginning as fun side project, Dynamic Prose Markup Language has turned into a viable and useful tool for our collaborative art works. We hope that with help and the additional profile of a Turbulence Commission, the project can be made available for other artists in their projects having ramifications beyond our knowledge.

Thanks for your time. If you have any questions, please let us know. Good luck with your decisions and we look forward to hearing from you…

Jeff Crouse and Steve Lambert

Our original idea grew out of a somewhat drunken conversation Dave and I were having about the site Last Night’s Party while walking home from Monkeytown one night. We both thought it was funny how people would go to parties just to get their picture taken, and how this artifact had become more important than the actual experience of the party. We thought it would be funny if there was a party in a huge, ugly, open space (like Eyebeam), with a bunch of movie sets where people could fake being at a really awesome party. And that would be the point of the party — to fake being at a better party. And thus, Anywhere But Here was born.

What we ended up with was a projector mounted on a pole, projecting images onto a wall. On top of the projector was a small video camera fitted with an infrared filter (which makes it see only infrared), and in front of the wall was a series of infrared panels, blasting the wall with infrared light. The idea is that, by blasting the wall with IR light, and having a camera that only sees IR light, anything that comes between the wall and the camera is seen as just a silhouette by the camera. So we built a little openFrameworks application that projects one of the background images full-screen and draws the silhouette seen by the camera on top in a solid color. The result is that the image being projected isn’t projected onto the people, only onto the wall. It’s kind of like a green screen, only much more ghetto – which is why we call it “The Poor Man’s Green Screen”. We basically stole the majority of setup from Zach Lieberman’s Messa di Voce – with his permission and help, of course.

The Backgrounds

Finding appropriate images was actually a lot harder than we thought. We figured out pretty early that it would be impossible to use most of the party photos we were finding online because, when projected onto the wall, the scale was all wrong. So first we had to find backgrounds with no people in them, then we had to crop and composite the people onto the images while projecting them onto the wall with someone standing up beside them. This is why you will see the half-legs and weird, ghost-like ankles and such. We instructed our photographer to aim high so that this stuff wouldn’t make it into the photos.

The Projection Software

This source code is an openFrameworks project, with project files for XCode and DevC++. It will expect you to have a webcam attached. I spent a lot of time on the menus because I wanted to be able to calibrate stuff on the fly – mostly because I was having so much trouble getting the silhouette to show up in the right place. The OpenCV can be set to do blob tracking or it can just paint the silhouettes on top of the image. I did this because I wanted to be able to give each blob an offset. It partially works.

There is also a paint brush feature to set the color of the blobs. I wanted to be able to control what color was projected onto the people so that the pictures would look more natural.

Unfortunately I only have footage of the software running on my MacBook, so it’s just looking at me with the MacBook webcam. The result is fuzzy blobs. As mentioned above, the ideal conditions involve using an IR lit wall and a camera with an IR gel.

awbh.zip

The Upload Software

We also had upload software created by the excellent Eric J. Olson. It takes the photos that our photographer shot and sends them to Facebook, Flickr, and any email addresses you put in. They are all captioned with something like “Check out this great party I am at! You should totally come!” Melissa has some great examples of what the software would upload to Flickr here and here.

I’m working on getting the source code for that and will post it soon.

The Result

Dave and I are currently working on a reel that will showcase the entire event, but in the meantime, here are all of the 2000+ photos that were taken that night. They were shot by Disco Meisch.

Overall, I was really impressed with how willing people were to really engage with the project. Nearly all of the 130+ people who got up on stage put on a great show – making it as entertaining for Dave and I as it (hopefully) was for them.

We will definitely be installing this piece in other locations, and we have some ideas on how to make the picture quality better, so we’re not done with this one yet.

While I am very pleased with how it came out and excited about how much people have liked it, it was, of course, never meant to act as a sincere step-by-step guide to anything, much less the complicated process of making a mixed reality project. So I thought I would write a little about some of the strategies we used in Invisible Threads and tell rest of the story – the stuff that was too boring or complicated to include in the video. Great intro, right?

The Approach

The more I work with installations and performances, the more I realize that people (at least the people who see my work) aren’t interested in technology; they are interested in an experience. Seems simple, but as a programmer, it is sometimes hard for me to remember. When I first began thinking about how to create a factory in Second Life, I immediately started planning out all of the systems and widgets that would work together to make the factory possible. These systems would be bigger than the project – they would be abstract enough to be useful to the rest of the world. But I quickly realized that this would be a pretty boring path to take. I would have definitely gotten bogged down in all of these separate systems, and I would have lost sight of the original goal. So what I ended up with was a pretty specific system with little that is going to be immediately useful to others. So none of this is going to save the world, but it gets the job done efficiently and I was able to concentrate on achieving exactly what I wanted, rather than being pulled in 5 other directions by the technological requirements I imposed on myself.

Anyway, on with the show. If you haven’t seen the Double Happiness factory yet, I recommend visiting it before reading on so that you have some idea of what I am talking about. <SLURL HERE>

The entire system is made up of 4 parts: (you will see by the end of this post that I enjoy spitting things up into logical parts)

1. the Second Life factory
2. the API/database service/Web Interface
3. the ExtendScript
4. the assets

1. The Second Life Factory

The first step is to have an input system in SL. For Invisible Threads, the system consisted of 2 parts: the machines and the template. For the sake of simplicity, we can say that the template is instantiated when one of the workers activates the loom (a specialized machine) by feeding it a prim that looks like a bale of cotton. The template looks like a flat rectangle with a jeans texture on it. When it is instantiated, a “start order” request is sent to the remote API (more about this later) and it gets back an order number. The template is scripted to move around the assembly line more or less on its own using this script:

Jean Template Movement Script

Each of the 9 machines is activated when a template passes over the machine’s “sensor”, an invisible prim that sits right in the middle of the machine. When this happens, the sensor tells the template to stop, tells the machine to turn on, and starts a timer. The sensor contains this script:

Machine Sensor Code

When the machine is activated by the sensor, the operator has N seconds to set the machine to the appropriate state based on the order from the customer. The operator at the machine clicks on the machine to change from one state to another. On each click, the machine enters a different state, which changes the front panel texture, and shouts out a message, like “size 34″ or “ragged hem”, or “Casual Friday syle” depending on the machine and the order.

Machine State Code

When the machine shouts out it’s message, the API script hears it, and when the timer ends, the machine makes a simple REST call using the llHTTPRequest() function. The call includes the order number (which was assigned by the API when the “start order” request was made), the name of the machine, and the last message that was shouted by the machine. The REST request might look something like this:

http://www.doublehappinessjeans.com/factory/index.php?orderId=34&machine=spool&message=hemmed

Machine API Client Code

And that is an over-simplified account of what happens in the factory! All of the Second Life scripting was done by my wonderful intern, Andrew Mahon, which is why the descriptions are a little vague and incomplete, but if you have any questions, please email me and I will try to answer them.

2. The Order Database Service/Web Interface

The web interface is made of 2 parts: the API endpoint and the Web Interface.

The API Endpoint

I tried to put as much of the business logic as possible on the server side because, frankly, Second Life scripting is a pain in the ass. The API endpoint is set up to distinguish calls from 3 different User Agents, ‘Second Life’, ‘ExtendScript’, and everything else. If it receives a call from anything but ‘Second Life’ or ‘ExtendScript’ (ie: a web browser), it will print a helpful HTML documentation page. If it receives a call from ‘Second Life’, it knows that the request is updating an order. We will get to the ExtendScript stuff later.

The important PHP classes are pretty straightforward: Order and Attribute. When the API receives a call from the machine called ‘loom’, it knows to create a new order. When it receives an order from the Quality Control machine, it knows that the order has made it all the way around the assembly line and it should be dealt with somehow. All other machines just send messages that get stored by the Order class in the order_attribute_lookup table. An Order object is really nothing more than an array of Attributes, an ID, a ‘status’, and a bunch of logic that tells it how to save itself and retrieve itself from the database. Later on, I started adding names to them, but that is just icing.

So when a request comes into the API that says “the dye vat just said that No Pants Left Behind was selected for order number 225″, the API will construct order number 225, add a new attribute, and save it back to the database. Of course, there is some sanity checks and there are the special cases that I mentioned above, but that is essentially all it does.

If the API is called by ExtendScript, it will check to see if there are any orders in the database whose status is set to ‘ready to print’. If there are, it returns a JSON encoded version of the order and then proceeds to put together the order, which I will talk about in a minute.

On literally the last day of coding, I realized that my mental model for this system was just a bit off. It would have been smarter to model the logic of the factory on what it actually was: a proxy for Photoshop. Instead of sending information about pairs of jeans, it should have been sending commands for Photoshop, like “Open Size34.pdf”, “Hide Layer ‘Roadkill’”, or “Print now”. I think this would have made the code more robust and easy to understand, and could have actually been used by other people. I don’t imagine I will ever want to take the time to re-write this server code, but if anyone else out there wants to make a factory in Second Life, that is what I would suggest

The Web Interface

While explaining Invisible Threads hundreds of times at Sundance and other venues, it occurred to the pragmatic scientist in me that the Double Happiness factory could be considered “the most inefficient way to fill out a web form ever invented.” That is, we went through all of the trouble building this factory, scripting these machines, hiring and training workers, buying an island, when really, what the whole thing is doing amounts to clicking a drop menu on an HTML form. Of course, the project was meant to provide an experience and to make people think, so the pragmatism of the system is kind of moot. But if the factory is the least efficient way to get this data into the database, the Web Interface is the most efficient.

It basically consists of one big table that contains all of the orders and an ‘edit’ button for each order so that you can change the attributes, status, and name of any order easily. This became an essential part of the project when, after weeks of training, I realized that the workers simply weren’t going to be able to consistently get orders right, even on the second or third try, no matter how much I trained them. So I paid a premium salary and the password to this interface to one worker whose job it was to monitor the orders and magically correct the orders if things weren’t going smoothly in the factory.

And that’s about it for the server-side stuff. Here is an archive of the code: Double Happiness Server Side Stuff

3. The ExtendScript

When we first started Invisible Threads, I was planning on doing all of the graphics generation with The Gimp and Script Foo. But since I am not a glutton for punishment, and not quite so strong in my Open Source principles that I would subject myself to that kind of a headache, after a few days of searching for the right Script-Foo interface and trying to pick up the arcane syntax, I gave up and discovered ExtendScript.

ExtendScript is the API provided by Adobe to control most Adobe products, such as Photoshop, InDesign and Illustrator. It comes in 3 flavors: JavaScript, VisualBasic, and AppleScript. It has a simple HTTP library which allowed me to query the database for unprinted orders and since I am pretty good with JavaScript, it was easy to get the job done. So basically, all the script does is run an infinite loop on a little Mac Mini that we hide away near the printer, polling the server for orders whose status is ‘ready to print’. When it finds one, it asks the server for a JSON version of the order, which it then uses to open the correct PSD, show and hide the appropriate layers, print the size information on a tag, and export the resulting graphic as a PDF file. The PDF file is then opened in Illustrator, and a message is sent back to the server to set the status of the order as ‘printed’.

Factory ExtendScript

The dirty secret of the project is that I never did figure out how to make Illustrator automatically print the generated PDF. I still have to go over to the Mac Mini and hit Apple+P. The settings to get the PDF to print properly on the large-format printer are so complex that I didn’t think it would be worth it to puzzle them all out. I *think* I could saved a preset and just told Illustrator to use it to print the document, but I kind of ran out of time.

4. The Assets

We decided to organize the jeans assets into 12 different PSDs by size. There’s not much to tell about them – they are pretty self-explanatory, so here they are. They’re huge.

• female, size 2
• female, size 6
• female, size 8
• female, size 10
• female, size 14
• female, size 16
• male, size 32
• male, size 34
• male, size 36
• male, size 40

Finally, one bit of icing that I was particularly fond of was the giant, 1984-style screen that was installed on one wall of the factory. I rented a QuickTime server from a very kind and courteous entrepreneur called Jamie Otis. His company, SLServer, rents audio and video streaming servers for $30/month, which is a very good price. Using QuickTime Broadcaster and an old iSight that I had, we were able to start a Quicktime video stream, and then it was just a matter of setting the media URL on the factory parcel to the URL of the stream, and that’s it. Instant live audio stream. This also made the employees happier, since they were able to see what was going on in the RL storefront.

1. Test Party 2: BoozBot and Murder Mystery – This Friday, the 30th, was our second Test Party. I will talk more about Test Party later, but suffice to say, we (Dave and I) are now committed to having a monthly party where we showcase the stuff we (and maybe some select others) are working on. This time around it was BoozBot (no documentation yet), a robotic bartender with remarkably lifelike conversation skills. He was – I won’t hesitate to say – a huge success. The Murder Mystery party was fun – dressing up and drinking with friends always is – but the actual mystery part was kind of lame. It was your typical murder mystery, called Four Deuces. I’ll give a full recount of it in a later post.
2. Teen Mashup Workshop at The Change You Want To See – This was a workshop organized by Andrew Patterson and Eyebeam, hosted by The Change You Want To See collective in Williamsburg. Dave and I were asked to come help a group of teens think about and build a party around a audio visual mashup performance they were going to develop one Saturday last month. Unfortunately, no teens showed up.
3. New Jeffcrouse.info site. I’ve updated my site to put more focus on the images and movies. I like it, but I don’t think it’s the most usable site in the world. I want to make an iPhone version – just to see what the iPhone development process is like.
4. Test Party 1: MadLips Karaoke and You’re So VIP – I admit, I was very scared that no one was going to come to this party. Half of the people who RSVP’d dropped out at the last minute, and we had like 4 people on the list. But then, somehow, like 30 people ended up coming by. MadLibs Karaoke was a hit, and even my hastily thrown together “You’re So VIP” – which was a sad shadow of what it was supposed to be: a 1-person VIP room where lucky individuals would be invited to go if they were cool enough. Once inside, the VIP would be praised by outsourced “praisers” in India.
5. Dirt Party @ Futuresonic – Manchester is a great town.
6. ABSML Turbulence grant – Yes, A BullShit Markup Language has received its first grant. Hopefully one of many.
7. Eyebeam Mixer: Anywhere But Here – We’re close to having a video about this one, so I’ll just leave it at that for now.
8. Inside the Artist’s Studio with James Chimpton – With Steve Lambert. The first (followed by BoozBot, of course) or what I hope will be a series of comedic, robotic “puppets” with varying levels of automation.
9. Invisible Threads @ Sundance – There is lots to say about this one. I’ll need a whole post. It was a pretty amazing experience.

I have to write about all of these things, and it’s a little intimidating. I think I’m going to have to sit myself in a cafe somewhere, put on headphones, and just write until I can write no more.

Unrelated Future Plans

1. I’m considering committing to adding one entry to Uncyclopedia every day for 100 days, similar to my 100 days, 100 stories project that I did in college.
2. What ever happened to Sunday Night Mixes? I should make a new one.
3. I wish I could arrange it so that, as I was walking down the street, people would be launching themselves into the windows of buildings from street level like a bunch of frogs. This kind of sounds like an Improv Everywhere project.
4. I wish I could pass a law that would make it illegal to have a job where your primary function is to make money by moving other peoples’ money around from one place to another. Seems this would solve lots of problems.
5. Travel Time (HopStop) Map – I dropped this idea at a random presentation to Greenpeace last Monday. The idea is that there is a huge map projected on the wall, and after specifying your travel preferences, you point to a place on the map, and the rest of the map would warp around the place where you pointed based on how long it would take to travel there. I will make this in OpenFrameworks
6. Overheard In New York Machinima Podcast – Something I’ve been wanting to do for a while, but I have had so little time to check on my Second Life island that my connections/rep there are probably all gone.
7. ReplicateMe.org
8. Inspired by Mr. Lambert’s collaboration with Prof. Duncombe, I want to write a book with a collaborator.

If you (my 5 friends who read this) are particularly interested in hearing more about any of these topics in particular, let me know and I will push them to the top of the list. I’m TOTALLY PUMPED to write more.

Do you want to earn Lindens by repeatedly pressing a button?

Do you want a small plot of land outside an industrial factory to build a hovel of your very own?

Are you available for 2-3 hours a day, from January 17th to January 27th?

Then contact me immediately! jobs@doublehappinessjeans.com

For the past few weeks, my life has been all about Invisible Threads, the project I am working on with Stephanie Rothenberg. We were recently informed that we will be presenting at a festival in January, so we’re kicking our asses to get it out the door and done. In a few words, we have built a sweatshop in Second Life that manufactures jeans. These jeans are then sent to a large-format printer and printed on Tyvek. They are then assembled and can be worn by the customer. So at this festival, we will be selling the jeans to anyone who has the cash.

Here’s the full project description:

“Invisible Threads” explores the growing intersection between labor, emerging virtual economies, and real life commodities through the creation of a designer jeans sweatshop in the online, 3-dimensional virtual world of Second Life.

Replicating a real life manufacturing facility that includes hiring Second Life “workers”, the project leverages the simulation capabilities of Second Life’s virtual space as well as its social networking tools to provide an insider’s view into current modes of global production. Low-wage workers are hired by placing classifieds in Second Life system and offering only slightly more than the countless other, extremely popular skill-less labor opportunities. At the start of each workday, workers clock-in. The worker is then be assigned to a specific department and workstation and given a specialized task to perform, all from the comfort of their respective homes or cyber-cafes all over the world. Ten machines, each correlating to a specification of the custom jean order, will be operated by a SL worker in an assembly line manner. For example, machine #2 which simulates a laser cutter creates the pattern whereas machine #3, a dye vat, creates the “rinse” effect. But just as in a real life factory, workers are monitored by a department supervisor and held accountable for their speed and efficiency and any production errors. The erratic flow of supply and demand and extenuating circumstances such as equipment failures and irrational dispositions may result in docked pay, layoffs and overtime.

Real-world customers place their order with the factory overseer, who manages the factory workers from a terminal in the physical space. As real world customers watch their jeans move down the assembly line via a projection on the wall of the physical space, the real lives behind the avatar “workers” stationed at industrial machines begin to emerge. This interplay between reality and virtual embodiment not only sheds light on the current politics of outsourced labor but foreshadows what has already become the future of capitalist production. At the end of the production cycle, the finished jeans are sent to a large format printer in the physical space, where they are printed on Tyvek material, quickly assembled, and worn out by the buyer. Styles include boot cut, skinny leg, flare and the new “Boyfriend” trousers. Profits from these purchases will be used to maintain the factory (monthly land rental tiers, SL advertising) and pay for workers’ land.

So I go to check my mail one last time before I go to sleep and what do I find?

Scary, right? Talk about nightmare-inducing. Talk about wishing I had LeechBlocked gmail after 1am.

But if you really want to look like ol’ Slambert, you need this.

[youtube]http://www.youtube.com/watch?v=-JAhth0RbhE[/youtube]

Make a long story short, Jeff.

Ok, Bocce got in, City Ball didn’t.

Bocce Drift is a no-tech variant of Bocce that Dave, Dan Klainbaum, John Goetzinger and I used to play back at Georgia Tech when we got tired of working on our theses. The essential difference between Bocce Drift and regular bocce is that a new round begins where the jack landed in the last round. This has the effect of moving a game around an area in an interesting way. Being on rough, unpredictable terrain, players must learn to use obstacles and geography to their advantage. And that’s about it.

So, we were a bit surprised that the conference organizers wanted us to fly to Amsterdam on our own dime to play a slightly modified version of bocce. But so be it. I want to go to Amsterdam, and this seems like a barely-passable excuse. In the meantime, Dave and I are going to put a little more effort into improving the game so that it is more interesting for us to create.