I Hired 8 Writers and Artists to Play a Game of Telephone on Fiverr

Fiverr is a service that allows you to easily find artists/writers/creators/whatevers to hire for very short contract gigs. Often times you can find someone for right around $5, hence the name Fiverr.

The Experiment

I used Fiverr a lot recently, especially with the artists I’m using for KnifeTank: The Shuffling, and I really loved it, so I thought it’d be fun to attempt an experiment. I wanted to see what would happen if I hired writers to describe an image, then hired artists to create images based on those descriptions. To start it off, I posted a photo I shot of Dance Party Robot along with this text “Describe this image in 300 words or less. Be specific enough that someone else could draw it based only on your description.”

Here is the text I received:

Background
It’s an image of a street side with a building in the background. The building has a deep blue colour. There are three large closed windows in the building that are green in colour. There’s a black coloured metal fence that covers two third of the background. The building has a modern look with square blocks.
Main Object
The main focus point of the picture is a Robotic figure. Its height is similar to the average human height. It has a square as a head, a triangle as the middle part of its body and rolled cylinder shaped legs. Its body is silver coloured that looks like aluminium. The hands are rolled sleeves made with aluminium foils. Its face has a laughing emoticon shape. There are four square black boxes at four sides and two small round black coloured shapes on the stomach part of his body.
Surroundings
We can see one-third of a street lamp by the side of the robotic figure. There are countless blurred pigeons, looking like they have just been scattered of all around the robotic figure. Only one pigeon is above the head of the robot which is flying desperately and beautifully.
Colour Patterns
The background is deep blue coloured. It’s past noon and the sun is still bright. The sun rays stop before the lamp and covers the whole robot and pigeons area.

I then asked an artist to create an image based solely on that description:

I sent that image to another writer on Fiverr and received this description:

Picturesque Description
This picture must have been designed with an Illustrator. This picture basically has a robot with a head like a square shape.
*The eyes are strategically designed in such a way that the head looks like a television.
*There are rays around the face like the blue sky.
*The robot itself is moving away from the direction of the birds behind it.
*The body of the robot looks like a rectangular shape at the upper part while from the chest downwards is more like a triangular shape without a sharp end.
*It has two skeleton hands; the ulna and radius are evidently showing. The part of the humerus has two bones like that of the lower arm but with a stripe of white dot joining the lower parts of the bones with the upper part. At the upper part of the arm is attached something like round ribbon with stripes across it.
*Additionally, across the chest are strips linked together with the upper arms’.
*The abdomen has like two small boxes as if they were windows.
Right under the small box to the left is a trace of the movement of the birds to the left of the building in front of which the robot is almost exiting.
*The path of the bird continues but bigger and closer to the floor than the one that pass through the robot at the left-back of it.
*The total color of the robot is grey.
*There is a big black bird at the back which is hauling up.
*The floor is tiled with bitumen like a road.
*Around the big bird are three other small ones diving to the left. There is a bird on top of the building close to the diving ones at the extreme.
*There is a house beside the robot with a street light.
*The color of the building is grey to the left and black at the other parts. Close to the robot are two rectangular shaped windows.
*These windows have green curtains with yellow shield.
*The Windows are barricaded with yellow color line round them.
*There is a door at the upper part of the house behind the robot to the left.
*This door has a read light on top of it.
*The sky is white and skyblue.
*The totality of the environment is cool and fascinating.

Which led another artist to draw this:

Which would later be described as:

IMAGE DESCRIPTION:
A robot with giant light blue button eyes stands there across the street with no expression on its face and its metallic body motionless. Its ears are what seems to be two antennas erected from a radio box except that these antennas are on the sides of its rectangular shaped metallic face. To add to it, a crescent-shaped metallic lip and thick rectangular metallic eyebrows complete its friendly looking face. However, it’s somewhat human-like body features two hands and legs that widens out on the lower part of them respectively, whilst a thin neck and lower body but a flat broad metallic chest complete the whole structure of this robot.
In the backdrop of this poised robot are the buildings with square windows on them and three green birds flying in the sky.

Which inspired an artist to draw this:

Things were going well until…

Things were going well until this point, but this is where the train quickly slipped off the rails. Despite hiring another writer for “technical description” of the previous image, I ended up receiving something a little more imaginative than I was expecting:

Humans are converted into robots. Because, robot has no feelings they only think about itself. They are destroying forest and making building there just for the sake of earning money and profit. According to them, they are making world beautiful and bright. But in real they through the world toward darkness.

I couldn’t imagine what the next artist would create based on that description, though I certainly wasn’t expecting them to create a two page comic:

Ending the experiment for ethical reasons

This is the point where I considered the project completed. I already felt like I was getting more work from my artists than I was paying for, but I couldn’t imagine trying to hire some poor Fiverr user to describe a poorly lit photo of some coloring pencil art so I decided to stop the experiment.

The results

I hired 8 contractors for about $19 each ($5 for the gig, $10 in tip, and $4 in Fiverr’s extremely high service fees), meaning I spent about $152 total. I had originally intended to publish the results in a 24 page zine, but I just can’t imagine how I’d throw this all together at this point, so I’m just sharing it here.

The artists were all great to work with and in most cases I think I received far better work than I paid for. Surprisingly all of my contractors completed their work on schedule and I only had one artist drop the ball on their work. There was a noticeable loss in data as the project continued, but it really was going better than I expected until the last two bits fell into place. That’s how these things go though and I have no regrets.

I wanted to give a huge shout out to all of the artists and writers who shared their skills with me, including:

 

Roast Beef Dunwich, our horrifying new Twitterbot

Have you ever wondered what items would be on the menu in a cosmic-horror-themed diner? Me too! Well not at first, but I had a fun Twitter handle (@eatroastbeef) that I wanted to do something neat with, so I thought it would be fun combine a bunch of unspeakable descriptions from Lovecraft novels along with tasty menu descriptions from various local restaurants and fast food chains. Ideally, the bot would then send tweets with a name, short description, price, and calorie count.  Like this:

Christine had written her own Markov Chain script for a similar project earlier this year, so we thought that would be perfect for this project. While Christine worked on the code, I collected as many descriptions as possible to feed to the bot. This was the most tedious part for me as I had to scour the web hunting down good sources, then usually needing to click copy and paste multiple times on each item on their menu and add them to our list. Many sites didn’t even have this info on one page, so I had to load a new page for every single item. This was hours of work, but it was so much fun to see the bots descriptions improve and change with each new menu we feed it. For the Lovecraft text, I started off by just feeding in all of his novels, but that often yielded some boring tweets. I was lucky to find Yog-Blogsoth, which collected every single Lovecraft creature (no matter how obscure) and listed it along with the specific block of text that describes it. This meant all of our Lovecraft text was just the gruesome descriptions, with none of that plot stuff to get in the way. It was still a ton of copying/pasting/checking for duplicates, but it saved a ton of time. 

We had some decent names to choose from, but ended up calling our “diner” Roast Beef Dunwich. I banged out a mascot and logo and the end result is online now. We don’t curate the feed, so there’s bound to be some gibberish, but so far we’ve seen some great tweets. Be sure to give @eatroastbeef a follow for a small dose of absurd tweets throughout the day. 

My favorite food title so far.
No wait, this is my favorite item title! “NYARLATHOTEP MASHED POTATOES THAT ENDED ARBYS”?! How did it even come up with that title? 
Gluten free!

Fun fact, we have yet to see a single Shoggoth or Cthulhu callout since our bot went live. I think I’ll freak out when it happens. 

For more information, check out Christine’s post about Roast Beef Dunwich… and be sure to follow and share our ghastly creation

 

Really Simple Style: How I RSS’d My Wardrobe

When a friend asked where I bought my shoes I told them it was the result of an RSS search on eBay. Looking closer, I realized my entire outfit (from head to toe) was the result of RSS searches.

Like many of you, I buy most of my clothes online. The only difference is that I used RSS feeds to find most of the clothes I bought in the last 5 years. Typically this means I’ll do a search for something on eBay or craigslist, and if they don’t have what I want, I add that search to my RSS reader so I’ll get notified whenever that item appears. Then I just forget about it and start watching cute animal videos on youtube.

From time to time an item appears in my RSS feed and I check to see if the Buy It Now is within my price range. If not, I come up with a low (but reasonable price) and use Auction Sniper to place my auction bid in the final seconds of a listing. Typically, I lose, but that’s a good thing because the item sold for more than I was willing to pay. The whole system is set up for me not to constantly check eBay for an item, then get into a bidding war on that item (that I don’t really NEED, but kind of want). As much as I love getting clothes for cheap, I also get competitive with other buyers and used to spend way too much, so this set-it and forget-it mentality really works for people like me that may be willing to have a search going for years before they score the perfect item for an insanely low price.

My RSS Outfit

Like I mentioned before, my entire outfit (except for the underwear) was found via RSS feeds. So we decided to do a fashion shoot documenting the hat, jacket, shirt, shoes, pants, and even socks. Huge thanks to my Emma Zwirko for taking these photos (and not making me feel super silly while doing it). You can see more of her fashion photos on Instagram.
An outfit comprised of RSS searches
… 

 

A Portrait Through a 100 Year Old Camera At ORDcamp in Chicago

Last month I flew up to Chicago for a 3 day un-conference called ORDcamp. I had a fantastic time and really appreciated all the extra touches they put into it. For instance, each attendee had their portrait taken by Pete Tsai, a Chicago based photographer. Pete had a 95yo Deardorff 8×10 camera set up, with a DSLR behind it. He’d have us pose in front of the Deardorff camera (with the shutter wide open, so you could see right through it), then he’d use the DSLR to take a photo of the camera with you behind it. Here’s the result:

OrdCamp Portrait

Despite being invited in 2008 and 2009, this was the first year I was ever actually able to afford the airfare and lodging for ORDcamp. Disposable income FTW! I’m so glad they invited me again, because it was an amazing experience. You know those people who go to Burning Man for the first time, then spend the next few months talking about it? That’s how my friends must have felt when I came back from ORDcamp this year.

Here are a few more of Pete’s portraits:

Taken with a 95 year old camera! #ordcamp

A post shared by Tanner Woodford (@tannerwoodford) on