One of the hottest tickets at the 80th Venice film festival isn’t a movie at all but a VR installation on the event’s self-styled “Immersive Island”. Each user sits at a computer and answers a series of personal questions, which the exhibit – in the space of a few seconds – converts into a bespoke portrait of their life. The project, Tulpamancer, is officially the work of Brooklyn-based artists Marc Da Costa and Matthew Niederhauser. In practice, though, it amounts to a creative collaboration between the user and AI.
Generative AI plays the role of Sleeping Beauty’s bad fairy at Venice. The ongoing writers and actors’ strike was largely prompted by fears over the new technology’s impact on film and TV production and has resulted in numerous star performers deciding to skip this year’s festival. But in the meantime, AI – unwelcome, uninvited and arguably misunderstood – has already joined the party. It’s hiding in the cracks of the films on the main programme and helping facilitate the creation of the XR (extended reality) pieces on the island.
For the organisers of the festival’s pioneering Venice Immersive section, the repercussions of the WGA and Sag-Aftra strike has brought some significant side benefits. Foot traffic to the island is up on previous years. Tickets to the exhibits have all sold out. “The strike has also helped us attract more focus from the press,” says co-curator Michel Reilhac. “That’s because there are less of the usual things that they focus on. Fewer stars on the red carpet mean journalists look elsewhere for their stories.” For an emergent artform, press coverage is crucial. It ensures that the experimental prototypes which are unveiled on the island can at least be read about outside Venice.
The downside is that Venice Immersive – with its XR experiments and VRChat tours – risks being tainted by fears over AI; viewed as the outrider for a potentially hazardous future. “People seem to think that XR is only about the technology and so they lump it all with AI,” says Liz Rosenthal, Reilhac’s co-curator. “Someone asked me yesterday, ‘Oh, are you getting lots of AI things funded now?’, which is a complete misunderstanding of what AI is. AI is just a workflow tool, like an editing or post-production tool. We have pieces here that use AI, but we’re not dependent on it. So there’s a big confusion around the issue.”
In Reilhac’s view, the XR community – which is still finding its range and seeking an effective model of distribution – remains altogether too small to be seen as a threat to the traditional entertainment industry. “There’s hardly any market [for XR work]. There’s no money in it. And I think that’s actually a beautiful feature of the community. It’s still a bit of a utopia, based on everyone helping each other and sharing ideas, because no one’s in it for the money. So I don’t think the studios associate us with AI. And even if they did, we’re not significant enough to be a threat.”
Elsewhere on the island I meet Paul Raphael, the co-founder of the immersive art company Felix & Paul Studios and co-creator of the mixed media animation Jim Henson’s The Storyteller: The Seven Ravens. Raphael explains that most XR creators are independent agents, non-unionised and still working on the experimental outer limits of the entertainment industry.
“Union laws and guidelines don’t really apply in the XR world,” he says. “The fact that it’s a new medium means it’s not regulated yet. But as we go forward, it will have to be. Right now the artform is small enough and young enough, and the distribution channels aren’t there yet, which means that it hasn’t broken through to reach a mass audience. As soon as it does, the situation will change. If the unions are able to survive the strikes, there is no question in my mind that the XR industry will eventually become a part of that structure.”
Raphael used predictive AI as a tool on The Seven Ravens in order to assist with the tracking of the exhibit’s digital animation. With Tulpamancer, though, the technology’s role is more central. Generative AI effectively mines the archive of information that it receives from each visitor. It digitally edits and illustrates your written responses and then delivers this back in the form of a dreamlike spiritual audit. “The yearning to explore distant lands beckons you, resonating deep with your being,” reads the computer print-out that I receive at the end of my own visit. “The call of Canada, Mexico, Argentina and Japan resonates alongside your family commitments and financial restraints.” I’m also assured that all the information I typed in at the start is deleted the instant the experience is complete
Niederhauser has spent the past week in Venice, ensuring the exhibit runs smoothly and observing the reactions of those who try it. Some come out tearful and joyous. Others, though, emerge disconcerted and disturbed. Such a breadth of response has made Tulpamancer one of the hits of this year’s festival. But it also reflects our own ambivalence towards generative AI, our fraught relationship with a technology that takes its lead from both the best and worst of humanity. The future will be interesting and does not necessarily spell disaster. “AI is not the ultimate good,” Niederhauser says. “But it’s not the ultimate evil either.”