No. 84
Alibabaum Sortium
By : Arvid Jense
Entrant’s location : Netherlands
LINKS
Description
(Multimedia installation: audio track and collection of paper objects. 2018) The Alibaba and Aliexpress search bars are the gateway to the thousands items sold by Chinese factories and resellers. Algorithms control when, how many times and in which order the searched products appear on visitors screens. By looking at the Ali platform like naturalists, we approached the digital platforms like ecosystems. We analysed several algorithms, narrowing in on the sorting algorithm and renaming it: Alibabaum Sortium. In the process of exploring this, we were drawn to specific products on Alibaba whose physical existence is a source of doubt: it is unclear whether they occur as purchasable items or if their role is to entertain customers and keep them browsing through Alibaba’s interface. Alibabaum Sortium is a collection of object simulations, in paper, of what Alibaba claims are its most desired products. The other side of the reconstructed objects is printed with the metadata that writes their existence into the digital world while we can hear the algorithm itself talking about it's job as a sorting program.
What did you create?
Alibabaum Sortium is a collection of reproductions of the most popular products on the Alibaba e-commerce platform. The live size paper objects represent how the sorting algorithm on Alibaba understands those products, partly through pixel images, partly through descriptions and metadata. Showing the algorithm's favorite products gives the observer an opportunity to understand the inner workings of the algorithm. A program that does not recommend what shoppers are likely to buy but rather strange clickbait object that keep shoppers entertained on the platform. Next to the objects, an audio track is streamed in the space: a fi interview with the algorithm itself revealing more of the research behind the functioning of these product curators on e-commerce platforms.
Why did you make it?
In each e-commerce platform, sorting algorithms curate what consumers will then be offered as a bed, a hat or a table. What motivations drives sorting algorithms to make these decisions and order the products in one way or the other? This process is invisible for the consumer even though it has a great influence on their choices, with this work we try to bring a better understanding of the inner workings of sorting algorithms in e-commerce. By looking at sorting algorithms like a naturalist would research an organism, we observe how in each e-commerce platform the algorithms develop a different behaviour. On Alibaba, the algorithm behaves most strangely, recommending products very little people buy: using a click-bait strategy to keep consumers longer on the platform.
How did you make it?
To find out about the algorithm 's inner working we conducted extensive research and interviews with the engineer of other e-commerce platforms. On the specific Alibaba programs we could access the scientific literature on their software. The mock up interview with the algorithm is written based on this research. The objects are made of light cardboard printed on in collaboration with Canon and an experimental flatbed printer. The objects were modeled using only information accessible and understandable by the algorithm itself including dimensions, images, etc.
Your entry’s specification
(Multimedia installation: audio track and collection of paper objects.) All objects are real size varying between 5cmx5cm to 3mx3mx70cm, the full installation fits in a 50m2 room but it can also be displayed with less objects. The audio track can be played on speakers or headphones.