Rebecca Skurnik
Rebecca Skurnik
aa-logo-01-01-01.png

Absurd Algorithms

Absurd Algorithms

 
 

project overview

For my thesis at ITP, I wrote a series of Absurd Algorithms as a satirical, social art piece. The piece portrays the reliance we have on technology, particularly the algorithms behind it that can potentially lead us awry.

THESIS TALK

 

idea

With the rise of technology, digital optimization, and algorithms, people have become reliant on systems that they do not fully understand. In addition, these algorithms are not perfect- some have biases, some confine us to options based on our categorization, etc. Despite all this, people still make decisions based on what algorithms have suggested to them. I then decided to be the “algorithm maker,” and build my own algorithms to make my own decisions. And I made the algorithms absurd— based off of interesting information that I found…because why not? For my thesis, I relied on them for a few weeks.

Try them yourself here!

 

THE ABSURD ALGORITHMS

algorithms.png
 

algorithm 1: finding a partner

Step 1

I went onto my dating app. I then took a screenshot of a guy’s picture and uploaded it the website I built.

interface.png

Step 2

Once my picture and my potential date pictures are uploaded, poseNet, a machine learning model in p5.js/javascript detects where my eyes and nose are on my face. I then did some math and took the slope of the line between both my left eye and my nose, and my right eye and my nose. If the result was similar enough to the same points on the man’s face, the algorithm would tell me to swipe right (say yes), and then if not, swipe left (say no).

posenet2-01.png

Step 3

Here are some examples from the algorithm. If it tells me that my facial expression matches the other person’s, then I swipe right (say yes) to him on my dating app, and if not, then I swipe left (say no) to him on the app.

notmatch.png

Step 4

After I swiping for a few weeks, I received some matches on my dating app- this means that those men who I said yes to, also said yes to me. We were then able to message each other. I spoke with some of the men I matched with, and went on a date with one of them…it did not go so well. The date was recorded and is shown in my thesis talk video.

 

algorithm 2: finding an apartment

I created a Google Chrome extension that scrapes my past orders from Seamless, and then tells me where I should live. The algorithm is based off of price, and the Seamless awards campaigns that I saw on the subway.

subwayseamless.png

I had some friends try this algorithm and see where it told them to live. It told one of my friends to live in Gramercy. It’s funny, because she only used Seamless when she lived in Brooklyn, and she had recently moved to Gramercy.

 

algorithm 1: finding a show to watch on netflix

Step 1

I researched Zodiac sign personality traits. I then found a website that has all the genre categories of shows on Netflix. I then matched each Zodiac sign to a genre that best matches their personality traits. Below is an example of a Libra and the genre I matched it with.

zodiac.001.jpeg

Step 2

I then plugged the genre number given from the website into the official Netflix site URL to get a list of shows under that genre and its thumbnail images.

net.001.jpeg

Step 4

I built a Google Chrome Extension that scrapes all of this information. I then created a JSON file with all of this information, and put it onto a website I created so that other people could try out the algorithm. Here is a video of the website.

Step 5

I had some friends try this algorithm and watch a few of the shows that the algorithm told them to watch based on their Zodiac sign. Their reactions were recorded and is shown in my thesis talk video.

 

ITP, Spring 2019

Thesis Class, Professor Gabe Barcia Colombo