Dreamcatcher: Social Media for the Subconscious

Visitors to our Dreamcatcher exhibition at Imagine RIT will step into a space where the boundaries between technology and the subconscious blur. Dreamcatcher is an RIT-student developed social media platform designed to let users share their dreams, explore their friends' dreamscapes, and uncover unexpected connections between them. The presentation would focus on our exploration of the development process and the technologies we've adopted to create Dreamcatcher, as well as showcasing the app and its functionality to the audience. This is a project we've built completely from scratch, so depending on the viewer's area of expertise we'd delve into our decision process for everything from database structure to aesthetic choices. Some of the visual aids we're planning to present to aid our explanation of what has gone into developing dreamcatcher include: 1. A snapshot of the Dreamcatcher database, including the embedded vectors we're using to represent dreams. 2. A flow representation of the different APIs and cloud providers we're using (i.e., how the app works behind the scenes). 3. The progress of the rolling whiteboard that's been the road map for our two-person developer team. 4. The actual app itself, available for download via QR code. After downloading the app or viewing our demo, visitors can enter their own dream descriptions through the app, the text of which will be instantly processed by a large language model (LLM) that transforms their text into vectorized data. This vector representation allows us to compare their dreams with others in our database, revealing similarities in themes, emotions, or even specific imagery. Visitors can explore a live feed of public dream entries and see how their dreams align with those of their friends, possibly even discovering eerie coincidences or shared dream motifs. For those interested in the technical side, we will have a behind-the-scenes look at how our AI-driven dream vectorization works, explaining the role of natural language processing, embeddings, and vector database access in making this analysis possible. We're interested in the potential applications of this data in social and psychological contexts, as it's exactly the type of dataset that would be impossible to amass without the ease and convenience of collecting it through a mobile app experience. Ultimately, we want to showcase Dreamcatcher's offering of a unique blend of technology, psychology, and creativity, allowing users to experience a new way of connecting with themselves and others through the world of dreams.