An innovative pop-up art installation exploring the realms of artificial intelligence (A.I.) and its potential for both destruction and good has concluded its initial run in San Francisco this month. However, the nonprofit organization behind the thought-provoking exhibit, the Misalignment Museum, has ambitious plans to establish the installation permanently.
The Misalignment Museum presented itself as an “art installation with the purpose of increasing knowledge about Artificial Intelligence” during its tenure in San Francisco’s Mission District, which began in March. While some select pieces remain on private display at the Salesforce Tower, the nonprofit is actively seeking funding to ensure its next public opening.
The project’s website states, “We hope to elevate public discourse and understanding of this powerful technology to inspire thoughtful collaboration, an appropriate regulatory environment, and progress towards a hopeful, vibrant future.” Among the captivating works showcased at the museum was “Spambots” by Neil Mendoza, a sculpture crafted using a Raspberry Pi, a credit card-sized single-board computer. The sculpture incorporates modified Spam cans with arms that control four keys on a keyboard.
The exhibition notes highlight that A.I. is increasingly being utilized to generate spam content, and the Spambots collectively type out prose generated by a deep learning-based large language model fine-tuned on a specially altered version of Aldous Huxley’s “Brave New World.”
Another remarkable piece, “Infinite Conversation” by Giacomo Miceli, employs machine learning to facilitate an eternal discussion between German film director Werner Herzog and Slovenian philosopher Slavo Žižek. A.I.-generated models trained on publicly available recordings of the two intellectuals enable this never-ending conversation. The gallery notes emphasize the project’s goal of raising awareness about the ease of synthesizing real voices using A.I., which carries significant implications for the media we consume and questions the importance of authoritative sources, trust breaches, and gullibility.
The Misalignment Museum also invited renowned Pier Group artists Kevan Christiaens, Hillary Clark, and Matthew Schultz, known for their monumental work at Burning Man, to create a sculpture for the exhibit. Their piece, “Paperclip Embrace,” constructed from over 15,000 paperclips and concrete, draws inspiration from the “Paperclip Maximizer” thought experiment on A.I. ethics, first proposed by Nick Bostrom in 2003.
Bostrom mused, “It also seems perfectly possible to have a superintelligence whose sole goal is something completely arbitrary, such as to manufacture as many paperclips as possible, and who would resist with all its might any attempt to alter this goal. For better or worse, artificial intellects need not share our human motivational tendencies.” Audrey Kim, the curator of the Misalignment Museum, explained that the project began around seven months ago, prior to the launch of ChatGPT, which sparked public concerns over chatbots and generative A.I. Kim intentionally structured the exhibit with dystopian pieces on the lower level and more optimistic works on the upper level.
Reflecting on the paradoxical nature of the subject matter, Kim expressed, “It’s weird because it’s such a terrifying topic, but it makes me happy people are interested.” The Misalignment Museum’s ongoing efforts to find a permanent home for its provocative exploration of A.I. signify a desire to continue stimulating dialogue and fostering a deeper understanding of this transformative technology and its societal implications.