AI-supported VR Project Immerses Audiences in Their Personalized Dreams
‘Tulpamancer’, a VR project debut at the Venice Immersive, creates personalized immersive narratives based on participants’ life experiences and dreams, using open-source AI.
The Venice Immersive, a sidebar to the renowned film festival focusing on virtual reality (VR), hosts a new art installation that turns the ethereal nature of dreams into tangible virtual experiences. ‘Tulpamancer’, a collaborative endeavor by Marc Da Costa and Matthew Niederhauser, is an art piece that uses open-source AI technologies and VR to recreate users' dreams and memories.
Speaking about the inspiration and intent behind the project, Da Costa revealed to The Hollywood Reporter Roma, “Tulpamancer is an interactive experience that allows people to have an intimate encounter with AI technology.”
The retro-themed setup exudes a distinctive 1980s vibe, with the Tulpamancer computer interface showcased on a classic black screen with fluorescent green text. The participants are invited to sit in a room in VR and answer questions about memories, past and future. The process is led by the enigmatic voice of ‘Tulpa’, a software program that, in the narrative, was found in an old warehouse once belonging to an East German scientist, and serves as the participants’ guide into their subconscious. The participants’ intimate interaction with the AI program and subsequent answers lead to the generation of new scripts, visuals and sound, constituting the immersive and personalized virtual narratives through a VR headset.
Niederhauser stated that the project was conceived with the initial artists’ perspective that regards AI as a new tool “to be displayed as an installation but also for storytelling”. The project was initiated before the recent Hollywood strikes that question the role of AI in art and storytelling, but 'Tulpamancer' certainly contributes to the conversation.
The creators of 'Tulpamancer' want the audiences to understand that while AI can be a powerful tool in storytelling, it has its boundaries. The installation's capacity to recreate memories and dreams is directly proportional to the details shared by the participant, and thus without sufficient input, the AI would struggle to produce a cohesive narrative, resulting in abstract and unclear renditions.
In this attempt to, in a way, “demystify” AI technologies in entertainment and art production, Niederhauser says that ‘Tulpamancer’ is only an installation but does foreshadow the coming possibility of personalized narratives and power to engage. Both artists acknowledge the potential harm brought by the technology, especially in the hands of exploitative industries, and emphasize the importance of regulation.