Readings 3/30/21:
Broken Promises and Empty Threats: From this reading, I’ve started to see how AI is typically funded by fear and retaliation, even today. American AI research today, part of a new boom built on economic fear is built to restore faith in the US economic system. Because of this, it seems to me that AI is built to protect capitol and punish people. AI works as Amazon HR reps now that actually can arbitrarily lower warehouse worker pay if it deems them not productive enough. Clearly, AI is currently going down a bad path, and I think we should start funding AI in more humanitarian fields where it could actually help rather than hurt, as it does in this reactionary space.
Enchanting Objects: From this reading, my main takeaway was how pertinent the idea that AI is a necessary good that requires some sacrifice to inbue in society is. AI, and AI design as it functions right now is very much in the hands of the few who know how to operate it, and it would appear no easy way of manipulating it is within public knowledge. I think AI is sort of gatekept with this idea that there is no way to really let people test the boundaries of what it can do, despite the fact that it becomes more possible every day just through the nature of what AI is.
Research Question: How can we the information gathering of AI to assist blue collar workers instead of harming them?
Initial Research:
My initial research patterns followed case studies of AI systems developed to help or harm workers, to see how AI was used to effect workers and their job security. I dove into Amazon’s predatory AI that watches and moderates worker payment, AI installed to “help” workers by increasing the amount of busywork they do every day, as well as studies done on the effects AI has on labor and the danger of automation that is always present.
Design Ideation:
My initial design concept was based in the idea of having a small activist robot that would travel through the city and accost businesses for treating their workers poorly. The robot would be dispatched from a center somewhere nearby and rush to the business to spout collected worker complaints. The complaints would be generated with an app, where workers could submit their complaints and the AI could determine how crucial help was, and would send the robot to act accordingly.
After showing this to Dina, the idea proved to be too complex and variable, and the idea of a more stable, immobile piece became a much more reasonable and interesting option. From this, I designed a few options for a dynamic art piece, a statue that would use the same type of AI to change its form and appear in different orientations based on the current state of the blue collar worker conditions in the area.
The AI was also adapted to be something that gathers its data from complaints on Twitter and other social media, allowing it to be more representative of local data as it draws from a larger platform.
The options I presented seemed limited, and loosely tied to the goal of the project. What did work, was their obstructive ability, as things got worse, the statues would become more and more obstructive to the people around them, with a hope that that would inspire change in society. This idea prompted my next iterations, which were built around this idea of discomfort bringing about change.
The next form my project took was a more interactive, personal level piece based more directly around Amazon itself, a major culprit of AI oppression and an important opponent for the fight for workers’ rights.
I drafted a box that would be set up as a street exhibit, prompting passersby to wear it on their head. Based on the same Algorithmic AI, it would make the experience inside the box worse depending on worker conditions in the immediate area. This exhibit was smaller and more personal, and took up less space in its area, meaning it could be installed near amazon stores or warehouses for a more relevant message.
Although, the box was a difficult interaction to start. Expecting someone to walk up to a box on the street and put their head into it was a big ask, and it was even more difficult to have them keep the box on their head if it was designed to be uncomfortable. I worried the interaction would leave no lasting message and would be discarded. I didn’t want to put it in the traditional exhibit setting though, as I worried it would lose its appeal to an actual cause. If people didn’t have easy access to it, it would not be able to accomplish its goal of inspiring activism in everyday users.
I followed this with a number of iterations trying to find a better way of opening the interaction, through a more surprising moment, or a more conventional entry point, and I eventually landed on the idea of a physical space that would be entered as a positive experience, but would be difficult and negative, and force people to live through the struggles of an Amazon warehouse worker.
I prototyped and designed a maze and the way it would work. It would be built from amazon boxes to show Amazon iconography and nail that image into the heads of the participants while they moved through the maze as its walls changed and shifted to trap them inside the maze. The maze walls would shift when the participants were nowhere near them.
Iteration:
The maze went from an interior piece to an exterior piece, placed in a park and designed to mislead people into thinking it was a fun Amazon experience instead of a grueling and exhausting experience.
I eventually made the decision to include a more dynamic presence involving the exit of the maze, with the exit becoming visible briefly, only to be obstructed by the moving walls, becoming even more frustrating to the participants.
I also decided to make the end of the exhibit a direct call to action. After leaving the maze, participants would be given information to help the cause and learn more moving forwards.