Augmented Reality Installation (2022)

Tech: Unity, C#, Python, Processing, Cinema 4D, Midjourney, DALL-E 2
Field: AI, Augmented Reality, Interactive Design

The Solus Project is a part of the M.A. thesis work titled Synthetic Realities and Hyper-personalized Media Content. The Solus Project exists in the form of an augmented reality application and installation that reflects and speculates on how the future of commercial personalization products might take advantage of extended reality and synthetic media generation. It also provides transparency on what kind of information and strategies are being used to predict our preferences by the industry today. The project uses psychological profiling and recommender systems known from e-commerce and social media platforms to predict the users’ preferences towards certain content to design and curate hyper-personalized user experiences using AI-generated content.

Personalization of media content has been used by many online platforms like Meta, Netflix, and Spotify for several years. The backbone of this is recommender systems, whose job is to predict what content their users prefer based on certain assigned objectives, e.g., keeping people engaged with the platform, maximizing clicks, etc.

One element in these systems is looking for users with similar preferences to whom to suggest similar content. However, some of the critiques these systems have received are that they control their users behaviour and emotions based on economic interests. Another point of critique has been that social bubbles are created by filtering the user's media content which is suspected to facilitate the creation of echo chambers and social polarization.

In recent times, psychological profiling has been used to further advance the tailoring of user-preferred media content and user targeting. The most well-known example of this is the Facebook Cambridge Analytica scandal, where this approach was used to influence voters in the USA 2016 presidential election with campaigns and media content tailored to their preferences.

Until now, we have mostly been exposed to human-created media content, but current development has broadened the usage of synthetic media such as deep fakes and AI image generation. This kind of digital content is mostly experienced in 2D on the screen through electronic devices such as computers and smartphones. However, with the growing accessibility to extended reality technology and Web3 lurking around the corner, a possible future way to experience this kind of content could be in more immersive synthetic 3D environments.

When the users are engaging with the Solus Project, they must first complete a questionnaire for the system to analyze and calculate their most dominant personality traits. This is based on the psychological trait theory known as the big five and includes the five personality traits; Extroversion, Agreeableness, Conscientiousness, Emotional Stability, and Openness to Experience. The user must additionally indicate their positive attitude towards a certain kind of media content by selecting the image they like the most from a collection of images.

The users are provided with a tablet running the Solus Project application that uses image tracking and plane detection to personalize the appearance of the generated reality. Within this reality, the result of the personality analysis is revealed, and AI-generated objects are placed. The users are additionally presented with suggestive shopping tips in line with their personality and media preferences.

Current technological advancement is making it possible to present users with an experience tailored to their preferences. How will these personal experience cocoons impact society when reducing shared cultural points of interest?

2022 Spektral-Raumohr, Berlin