I started this project with two concepts in mind. The first was looking at fashion photographer Terry Richardson and trying to bring to light his inappropriate actions. He is known for sexually harassing the models he works with. My original idea for the Twitter bot was to retweet anything he tweeted and post it to my page with a link to an image that I created.The image is shown below:
Many of these sorts of images are easily found online once googling his name. I just manipulated it and put a filter on it so it wasn’t so vulgar (also it still is). I set this Twitter bot up and ran into a few issues. First, it was posting to my personal twitter account rather than the one I set up, which was an easy fix. Second, the image was somewhat hidden within the tweet. The link to his photograph that he put in the tweet was more so visible than mine, and it looked as if I had just retweeted his tweet with no other fixes. Here is a screen shot of his post without my link:
Unfortunately I got frustrated and deleted the posts before I got to document them which I realize would have been helpful now. My link was smaller and unnoticeable, therefore I figured more people would have clicked on the first link they saw, which was his photo and not my manipulated one. I would’ve liked if my image showed up directly along with his post, but I realize Twitter does not work like that.
For the programming aspect, I used IFTTT in order to get this up and running which was convenient and easy. My recipe looked like this:
Although this was a great attempt, I wanted to try something else and see where it got me.
This is when I turned to my other idea, about Oblique Strategies. What I was inspired by was text and language and the exercises in which we manipulated a source text to get a completely different result. For instance, we took poems and reversed the words or took out every other. I think that really fascinating things happen when language is experimented with and I wanted to explore this further. What really interests me is how words lose or change meaning once placed in a different order or repeated. I wanted to work with this and see how I can manipulate language and break it apart.
After looking through the archive of Oblique Strategies on Twitter, some of these strategies even motivated me to do this exercise by what they were saying.
I thought to myself, yes they do need changing. And that’s why I wanted to create this Twitter bot. Or things like this:
This Twitter bot was going to be all about emphasizing the flaws and removing specifics. Along with this, I realized that I enjoyed the Oblique Strategies more when they were random words put together. For example:
For some reason, that inspires me a lot more than actual sentences or questions. This is when I decided that I wanted to create a Twitter bot that takes random words from the Oblique Strategies tweets and creates new phrases and orders. I used a Python library called TextBlob in order to do this. What it does is provide an API to pick apart tasks dealing with language, “part-of-speech tagging, noun phrase extraction, sentiment analysis, classification, translation, and more.” I wrote a function in the code to take the random text and generate something new. I used Sublime Text in order to do this, typed the Consumer Keys, Access Keys and then also used Tweepy to assist which allows communication to Twitter.
TextBlob <–Website Here
Here is part of the code.
This generates new text from the existing Oblique Strategies.
This is the Twitter page:
And here are some examples of what it has generated so far:
Oblique Oblique <– Twitter Bot Link
Right now it is just running off of my computer but ideally I would have liked to put it on a server so that it is automatic and runs more often. I have been experimenting with Heroku multiple times but it is still causing issues with the git command, as we figured out it has something to do with my operating system.
Although the random effect is so random that it repeats a lot of the same words so nothing is clearly legible. Originally I didn’t like that it was hard to understand but now I like that it is so “bot” like because it goes with my goal of breaking apart language. I feel that this is more of a work in progress than a finished piece, but I feel that I am onto something that could eventually work well.