A new art installation at Venice Immersive – the film festival’s VR sidebar – promises to turn dreams into reality, at least a virtual reality.
Tulpamancercreated by Marc Da Costa and Matthew Niederhauser, leverages open-source AI tools including Open AI and StableDiffusion to reconstruct users’ memories and dreams in a guided two-step journey and then reproduce them through virtual reality (VR).
In the trailer of Tulpamancerwe hear the voice of “Tulpa”, a software program that, according to the play’s story, was discovered in an old warehouse that once housed the laboratory of an East German scientist.
“Tulpamancer is an interactive experience that allows people to have an intimate encounter with AI technology,” says Da Costa in an interview with THR Roma. “You walk into a room, sit down at a computer and chat with this mysterious AI entity that asks you questions about your memories, your past, your future. Then, with a VR headset, you enter a room where an entirely new, custom script, footage, and voice-over is produced for you.”
The style of the installation is reminiscent of the 1980s, with the tulpamancer computer displayed on a classic black screen with green text. According to the creators, the experience is an experiment on two fronts. The first is about the use of AI in art to contribute to the ongoing debate, which has culminated with the double whammy in Hollywood. The second is an experiment in constructing personalized narratives.
“We conceived this project before the strikes and our initial perspective as artists is that AI is a new tool that can be shown as an installation but also for storytelling,” says Niederhauser. “It’s a technology we’ve both been interested in for decades. Obviously, there has been a sudden shift in the quality of AI and its use, and there will definitely be abuse,” he continues. “We support the strikes and want to bring some differentiation into the discussion. In the hands of an exploitative industry, AI can be very harmful.
“We’ve created something that hopefully people can see [AI’s] Limitations, we didn’t create the matrix, it’s just an installation,” notes Niederhauser. “But also a foreshadowing of something that’s definitely to come, and it’s all about personalized narratives and the ability to engage and have a conversation.”
Testing the installation at Venice Immersive, it is clear what the two artists want to convey: as storytellers Tulpamancer is limited. The more detail shared with the machine, the more accurate the VR creation becomes, but otherwise it doesn’t have the tools to recreate anything at all, leaving details and imagery vague.
Despite TulpamancerGiven the harmless appearance, Da Costa warns that AI technology is “really powerful and needs to be regulated.”
Check out the exclusive trailer for Tulpamancer THR Roma‘S site here.