Taste Testing
Working with AI, not having it work for me.
Over the past couple of years, AI has seen a massive increase in popularity and in the mediums that it's available in. It's impossible to deny the scope of AI and its implications; despite its reach, it's unwise to disregard the arguably unethical practices that are the foundation upon which these algorithms exist.
Over the span of a month, I created a machine learning algorithm that serves, describes, or comments upon an aspect of my everyday life: the design process.
I trained an AI model using my own work, as well as the submissions of consenting peers, and embedded it into a site that offers an assessment of submitted pieces on the basis of what I considered to be design that suited my tastes.
Over the past couple of years, AI has seen a massive increase in popularity and in the mediums that it's available in. It's impossible to deny the scope of AI and its implications; despite its reach, it's unwise to disregard the arguably unethical practices that are the foundation upon which these algorithms exist.
Understanding the Role of AI Here and Now
Research
Before experimenting with my algorithm, I needed to understand how AI models worked and some core ideas behind artificial intelligence as a whole. I wanted to approach this project with a neutral stance, neither hypercritical nor extolling. I also looked into already existing AI models that in some way emulate elements that I want to incorporate in
my project.
Learning about the processes that power AI only drove me to further emphasize my need for ethical data collection for this project. My biggest question was “Do I want to be a good designer, or a good person?” I was determined to achieve both.
Things I liked…
And things I didn’t.
The Process
Harvesting + Sorting Data
I created a Google Form in order to gather responses from my peers. I made sure to use very precise language so that they understood why I wanted their input, and where that material was going.
Once I gathered my data, I sorted it into things I like, and things I don’t. I ended up putting some of my old work in the mix as well. Seeing my style develop through time was really interesting and rewarding.
