Ruoxi Wang is a graphic designer based in London and Beijing. She graduated from Central University of Finance and Economics, China, BA Visual Communication in 2018.
Awards：
Silver Award, CGDA Graphic Design Academy Award, 2017
100 Design of 2020, Award360, 2020
Ruoxi is passionate about digital art. She uses digital techniques and algorithms to explore the interaction between humans and technology and attempts to visualise this relationship in a series of digital graphics and animations.
Humans are living in many computational models, large and small, especially in networks. Every time we interact with a network, it is recorded as a learning resource for a particular machine. People make decisions within the digital realm that are not independent, but that are made by humans in collaboration with machine intelligence. In this collaborative relationship, we change our habits and behaviours because of the convenience or new experiences that AI brings, and AI learns how to interact with humans on a broader scale because it gains more data on human behaviour. Ruoxi explores this process of interplay and presents it as a series of digital works.
An Ego of Instagram is a one-minute animation and a website for presenting ideas and processes. It takes the audience inside the brain of the AI in the network and watches the process of the AI interacting with the user. It's a journey that shows the process of interacting with various AI software through sorting, searching, and composing, starting with people getting new information on Instagram. The process is also a reflection of people looking back on themselves, on what artificial intelligence in the web has brought to them, and what it has changed.
The project starts with Instagram, where the intelligent software's analysis, predictions and other actions on the user all take place in the computer that cannot be seen. The demonstration is a journey to explore the interaction between AI and users in the web. Google Photo, Google Image Search and Artbreeder are used in the work to represent the smart software that people come into contact with in their lives, and the videos generated through these software to represent the changes that occur when people use them. These changes can be positive or negative, dramatic or minimal. But what has to be acknowledged is that smart products have already had some kind of impact on humans. An Ego of Instagram tells the story of this impact from a perspective that allows people to enhance their understanding of their relationship with artificial intelligence.
Instagram has many popular functions such as explore, Instagram Stories, IGTV and Instagram reels. All these functions are born on the basis of algorithms. Instagram algorithm dictates the order of the posts that users see when they’re scrolling through their feed. Based on specific signals, it prioritises the most relevant posts by pushing them to the top and giving them maximum visibility. These are the three factors that influence the content in the feed.
The first one is Relationship with the user. If a certain user has interacted with a lot of your past Instagram content, they’ll be more likely to see your future content.
The next is the Interest the user has conveyed. This signal is based on whether the user interacts with similar posts and accounts when they explore Instagram. Users who also engage with similar content are more likely to see their own posts.
The third one is Relevancy of the post. When you publish a post on Instagram, the algorithm gives it a relevancy score, which impacts who it shows in the feed.
Therefore, I have designed an experiment to observe the characteristics of the ‘signal'. Only click and read the content of the explore page without additional interaction, such as no search, no likes and no comments. In order to make the results as independent of human behaviour as possible. I set up a new Instagram account for the experiment and turned off cookies and location. In the experiment, I refreshed the page after view every 20 posts, for a total of 25 refreshes and 500 posts. The 500 posts represent the information that users get when using Instagram. Each refresh means that Instagram has made a calculation of user preferences. The experiment simulates the process of mutual learning between the user and the computing system, where the user views new information and the computing system gets the resources to learn the person's preferences.
What kind of content would the system prefer to push to the account without any 'signal' from the user? Thus, I sorted the collection of 500 posts. I used Google photos, which is also based on an algorithm, to classify all the posts.
I wanted the result of this work to feel like a journey, where the viewer reflects on themselves while following the footsteps of the Instagram algorithm as it tours the development of an account. So I used the image compositing software called Artbreeder to the images from the collection of posts into a one-minute animation. The animation is played in sequence according to the order in which they were collected from Instagram. The animation conects the content of the posts together so that the audience can see the gradual change of the scenery around them as if they were on a tour through the posts.
At the same time, I was trying to understand what the images in the posts were saying. I therefore used Google image Search to search the collection of posts. I hope the audience will join me in exploring these posts. So I used blender to import the 3d models of the identified locations and stitched them together. And use the camera to move through the 'posts', walking through each of the locations in the posts.
