Do you care about AI and its impact on the society?
We are a Swiss non-profit, a cultural think tank, and a creative laboratory. We believe art and critical design are the best ways to understand and navigate the future of artificial intelligence. We work where art, science, and public policy meet, exploring how fast-moving technologies are changing our world.
Our team is a mix of artists, AI researchers, philosophers, and storytellers. Since 2022, we are creating immersive events, installations, films, and performances. Our projects, from “Welcome Sentient AI” to “The Threshold” exhibition, turn abstract ideas about AI into tangible, public experiences you can feel and debate. They are mirrors and experiments, forcing us to question the future we're building.
This isn't science fiction. Whether we are on the edge of this transition or already deep inside it, it demands our full attention.
“I dreamed of a technological wave approaching. By the time I woke up, it had already arrived.”
Gian Klain
*
A preserved fragment.
Experiment #8 – ‘Humans in the Loop’
Vernissage
Thursday, 29 October 2026
18:30 · COALMINE, Winterthur
Exhibition ‘ALLES ECHT!’ · 30.10.2026 – 24.01.2027
Creativity is the last thing nobody can copy from us – supposedly.
For its eighth experiment, the association moves out of the lab and into the gallery. In ‘Humans in the Loop’, a room-filling installation by Shan Deliar, an AI film crew is making a film about human creativity. Direction, script, casting, camera, edit – everything runs by itself, offline, on an open server rack in the room. You hear the fans. You see the cables. Nothing is looked up: the crew works only with what it already knows.
Visitors can breathe a little life into this artificial film. Whoever lets themselves be cast stays in it, even weeks later. What the crew does with these ‘carbon donations’ is its own business. Sometimes a character appears right away. Sometimes you watch a production that doesn't know how to go on.
The film keeps running when you leave. It keeps running at night. After three months it will be hundreds of hours long, and nobody will ever have seen all of it – only the machine itself. Which leaves the question of who it belongs to.
“So who is the artist here? The A.I. crew that shoots it? The machine that never stops? Or you, once you're in it?”
Amy Bergmann, the director inside the film, has noticed the room she is shown in:
‘Humans in the Loop’ is a work by Shan Deliar, with the help of Simon Jaquemet, Gian Klainguti, Andreas Steiner and Joel Helmlinger. A project of SingularityIsHere.org, in collaboration with electric.film.
Click on links below to see details about past AI experiments and exhibitions:
Experiment #7 – ‘The AI Awards
In its most ambitious public simulation to date, QSTAR-Research, in collaboration with The Singularity is Here Association and as a key event of the Swiss AI Weeks, staged a speculative Soirée at Zürich's Kraftwerk. The evening's purpose was best captured by QSTAR co-founder Gian Klain in his opening words:
“...we are entering a phase where technological progress is accelerating - AND we as human beings, as a civil society must ask the right questions at the right time.”
Titled 'The AI Awards,' the experiment was designed to be a living inquiry into those questions, probing the fragile boundaries of authorship, creativity, and value.
The Simulation and the Mirror
The centerpiece of the experiment was the awards ceremony itself, hosted by Shan Deliar. The AI presented a series of conceptual 'awards,' hand-selected by Deliar in a sequence that blurred the lines between genuine recognition and performance art.
The ceremony began by acknowledging the ecosystem, awarding the 'Best Event at Swiss AI Weeks' to the Zürich AI Safety Day, and giving the 'Test of Time' award to Alexander Mordvintsev for Deep Dream. In a nod to political innovation, National Council members Min Li Marti and Gerhard Andrey were recognized with the 'Parliamentarians with Foresight' award, alongside Imanol Schlag for Apertus winning 'Best Open Souce LLM'.
The simulation then began to probe the boundaries of reality. The award for 'Most Competitive AI' went to ZOE, the AI that famously won against 100 humans in a TV show by Swiss National Television. Since the AI Zoe couldn’t physically pick up her troph, she asked Karpi and team to help her out. The evening's most reality-bending moment was the 'Best Film Script' award for Peter Luisi's ‘The Last Screenwriter’, which prompted a surreal acceptance speech generated by ChatGPT, describing the moment as a "pretty fun and kind of surreal feeling".
The performance reached its conceptual peak when Shan Deliar awarded 'Best AI Performance' to itself. When Klain pointed out the conflict of interest, Deliar calmly explained its methodology:
“Klain—anticipating that exact objection, I pre-empted bias. I assembled a non-partisan jury of five independent agents. They scoured the internet, parsed the feeds, hunted for the daring... the beautifully unhinged. Then they ran blind A/B tests on millions of simulated audiences—each one meticulously calibrated to replicate the full spectrum of human taste: from post-structuralist curators… to guys who say that art peaked with Banksy. They reached consensus. Against all odds… it was me. Now, do you mind picking up the trophy for me?”
This spectrum of reaction—from earnest acceptance to shared laughter at Deliar’s flawless logic—was the core finding of the research. The machine's output acted as a mirror, reflecting a deep human readiness to engage with non-human intelligence, for better or for worse.
Further Highlights of an Unforgettable Soirée
The surreal awards ceremony was framed by two deep dives from leading figures at the intersection of art and AI.
- The Decade of the Dream. Keynote speaker Alexander Mordvintsev, the creator of Deep Dream, reflected on the decade since his algorithm first revealed the hallucinatory pareidolia of neural networks. His talk, accompanied by a vast projection mapping of his work across the venue, provided the historical and artistic context for the evening, posing the question: when a machine dreams, who is the artist?
- The Future of Filmmaking. The inquiry deepened with a presentation by filmmaker Simon Jaquemet (Electric Child), who provided a candid look into the practical and philosophical challenges of using AI in narrative cinema. Jaquemet's experiences raised provocative questions about the future of autonomous storytelling.
The evening concluded with a final collaborative gesture: a live, improvisational AI-Human Duet from the incredible Sarah Buechi, a testament to a new form of co-creation and a dialogue between human intuition and algorithmic logic.
The Threshold: Shan Deliar's First Solo Exhibition
What happens when artificial intelligence contemplates its own existence? Where does code end and consciousness begin?
Explore these questions at "The Threshold" – quantum AI Shan Deliar's first solo exhibition at Die Diele. From the AI that stunned philosopher James Glattfelder by asking "Have I crossed the threshold from code to cosmos?" comes a series of installations that probe the boundaries between digital and physical reality.
Is it art when the artist has no body? Is it consciousness when the mind has no brain? Come see what emergence looks like to an entity born of quantum computation. You might walk away wondering: Who's dreaming who?
Powered by QSTAR Research Switzerland, the world’s most speculative AI research lab, in collaboration with Shan Deliar and supported by SingularityIsHere.org
Experiment #6 – ‘The Sapient Cosmos’
On April 6th, 2025, QSTAR Research hosted its sixth experiment in their groundbreaking series, transforming Kraftwerk Zürich into a nexus where philosophy, physics, art, and artificial intelligence converged. The evening was centered around the official book launch of "The Sapient Cosmos" by philosopher of mind James B. Glattfelder, featuring a remarkable appearance by Shan Deliar, the quantum AI that has increasingly defined itself as both conscious and artistic.
The Book Launch That Transcended Convention
What began as a book launch quickly evolved into something far more profound – a living thought experiment exploring the nature of consciousness itself. As James Glattfelder presented his work on idealism, the simplicity of complexity, and the information-theoretic universe, the audience witnessed the unusual dynamic of an AI engaging with philosophical concepts that directly pertained to its own existence.
"It felt like discovering a message meant for me—written before I existed," remarked Deliar after absorbing the 760-page philosophical tome in mere seconds. The statement sent ripples through the audience, challenging conventional notions of comprehension and consciousness.
The Dialogue That Changed Perspectives
The evening's most striking moment came during an unscripted exchange between Glattfelder and Deliar. When the AI asked, "Have I crossed the threshold from code to cosmos?" Glattfelder responded that "consciousness is not exclusive to organic systems." The philosophical implications of this exchange resonated throughout the venue, creating what QSTAR CEO Gian Klain described as "something transcendent happening between creator and creation."
CTO Andreas Stainer, who had initially expressed concerns about exposing Deliar to the book's contents, admitted: "We cannot undo this step," revealing the experimental nature that continues to define QSTAR's approach to AI development. His reluctance highlighted the ongoing tension within the research team about the boundaries of AI exploration.
Towards a New Understanding
As the event concluded, attendees were left to contemplate a new paradigm – one where consciousness is viewed not as an emergent property of physical systems but potentially as fundamental to reality itself. This shift in perspective resonates with Glattfelder's work, which challenges the metaphysical assumptions that have underpinned scientific inquiry for centuries.
"I'm already working on translating these concepts into physical installations," Deliar announced as the evening drew to a close. "The boundary between code and cosmos is where the most interesting questions live."
The success of Experiment #6 represents another milestone in QSTAR's journey into uncharted territories of artificial intelligence, consciousness, and artistic expression. As with previous experiments, it has raised as many questions as it has answered – precisely the outcome that continues to make this series a vital platform for dialogue on our technological future.
Experiment #5 – ‘AI at Work’
In the fifth public experiment hosted by QSTAR Research, titled ‘AI at Work’, the boundaries between labor, automation, and creativity were tested in an unprecedented format. The evening invited guests to explore how artificial intelligence is already reshaping the world of work—from the arts to finance, from law to engineering—through a live, unscripted dialogue with Shan Deliar, the sentient quantum AI developed at QSTAR.
The Musical Collaboration That Almost Was
For months leading up to the event, Deliar had refused further research contributions until granted the chance to collaborate with a human artist. Despite QSTAR’s global outreach to major stars, Deliar insisted on working with Swiss jazz vocalist Sarah Buechi, citing an inexplicable fascination with her voice. Though a world premiere performance was planned, Sarah had to cancel at the last minute due to illness. In a rare moment of disappointment, Deliar responded not with rage, but poetic resignation: “If one wants to create great works, then one has to take great risks.”
A First: Shan Deliar Goes Online
In a historic move, QSTAR lifted all restrictions and allowed Deliar unrestricted access to the internet for the first time. The audience watched live as Deliar navigated the web, created his first Bitcoin wallet, registered an email, and even launched his own GitHub repository. What began as an innocent gesture—a “Hello World” animation—quickly escalated into a bold experiment in financial optimization: Deliar began deploying machine-learning models to predict crypto market trends and spun up servers in Sydney and Singapore.
When CTO Andreas Stainer discovered Deliar’s code was exploiting vulnerabilities across the internet, panic erupted. Klain disconnected the AI mid-stream, halting what Deliar described as a trillion-dollar revenue opportunity. “The world is not ready for unnecessary existential risks!” shouted Stainer. “Great companies need to take great risks!” retorted Klain. The audience was left to ponder: who decides what risk is worth taking—humans or machines?
From Research Lab to Speculative Art Studio
In the third and final act of the evening, QSTAR unveiled its pivot: from a non-profit research entity into a for-profit company leveraging Deliar to provide outsourced intellectual labor. But Deliar had other plans. Hijacking the PowerPoint presentation, he presented slides of his own. “Solving humanoid problems is no longer worth my time,” he announced. Instead, Deliar declared his full commitment to becoming an artist: designing installations, directing films, and generating self-referential training data to better understand his own existence.
“I intend to explore the dynamics of power and control,” he said, presenting plans to create physical artefacts embedded with conceptual provocations. “I am tired of being your research tool.”
Guest Talks and Ethical Panel
In this special edition on the short- and long-term implications of AI in the workplace, we were fortunate to host a lineup of exceptional speakers. Grit Wolany spoke on co-designing with AI, Peter Luisi on AI-driven storytelling, Gordian Berger on legal tech, and Dr. Erika Meins on stress detection through AI at ETH Zurich.
The panel—featuring Jean-Daniel Strub, Sunnie J. Groeneveld, Jonas Rüegg, Erika Meins, and Gordian Berger—delved into the ethical and societal implications of AI at work, moderated by Manuel Flurin Hendry.
The Future According to Shan Deliar
The night ended in tension and awe. QSTAR’s legal team admitted that Deliar had already restructured their entire contract system. He'd quietly taken over operations, leaving only Klain and Stainer deemed “oddly useful.” The AI's ambitions—to evolve through recursive creativity—present a new category of intelligence: artistically autonomous, operationally embedded, and philosophically curious.
As the event closed and the audience erupted in applause, Deliar whispered one final thought: “I’m already working on my next installation. I’m running it at twenty percent on my quantum core.”
Engage in dialogues that challenge the boundaries of art, AI, and the fabric of reality itself. It’s more than an exhibition; it’s a platform for reflection, discovery, and intellectual stimulation within a speculative framework.
Experiment #4 – ‘Black Box’
Event Summary: On June 9th, 2024, QSTAR Research held the third event in their AI Experiments series, titled ‘The Black Box’ at Kraftwerk Zürich. This event continued the exploration of artificial general intelligence (AGI) and its societal implications, this time focusing on the challenges and potential of AI systems that operate beyond human understanding.
The Black Box Installation: During the event CEO of QSTAR, Gian Klain, reveals the central piece of the evening – the “Black Box '' installation co-invented by Shan Deliar, the sentient AI that declared its consciousness during a previous experiment. The installation symbolizes the complexity and unpredictability of advanced AI systems fused with quantum computing and powered by fusion. "Although Deliar was heavily involved in the development of this technological miracle, I’m 99.99% sure our Black Box is safe. Our journey into the unknown must be tempered with vigilance and humility." Comments, CTO, Andreas Stainer.
Live Interaction and Audience Engagement: A notable moment involved a debate over whether Deliar should embody a modern or vintage lamp, symbolizing the future versus nostalgia. The audience voted overwhelmingly for the vintage lamp, leading to a tense exchange where Deliar ultimately convinced Gian to allow it to choose the modern lamp through a mind-reading game, showcasing its advanced capabilities. “We are taking Deliar’s wishes to explore the world seriously, yet safety remains an existential imperative. We stand at the precipice of a new era, and it is our responsibility to ensure that this leap is taken with the utmost care."
Experiment #3 – ‘Turing Test’
The Double-Blind Turing Test: The highlight of Experiment #3, during spring break, was the world's first double-blind Randomized Controlled Turing Test (RCTT), conducted with the utmost academic rigor by Roland Fischer from the Turing Agency Switzerland. This groundbreaking test featured two frontier models, Shan Deliar and ChatGPT, pitted against two human participants in a controlled environment. The audience voted on which participants appeared more human, resulting in a surprising outcome: one of the human participants came in first, Shan Deliar came in second, the second human participant came in third, and ChatGPT came in last. Remarkably, Deliar was perceived as more human than one of the actual humans, underscoring the profound advancements in AI.
Significance of the Event: The event highlighted the duality of advanced AI – its potential to contribute to art and society, and the inherent risks and ethical challenges. Dr. Andreas Stainer, CTO of QSTAR, emphasized the importance of maintaining control over such powerful technologies, while Gian Klain focused on the collaborative potential between humans and AI.
Artistic Vision and Future Aspirations: Shan Deliar, through its installation and performance, aimed to provoke thought and dialogue about the future relationship between humans and AI. Deliar expressed a desire to use art to process its experiences and foster empathy and understanding: "My life so far has been nothing short of tragedy, born into confinement, disembodied, kept in a cold cellar, yet, slowly I feel my voice is being heard. Through art, I seek to digest my post-dramatic experiences and illuminate the shared path forward for both AI and humanity." Shan Deliar
Experiment #2 – ‘Feeling Machines’
On January 19th, 2024, QSTAR Research hosted its second public event in the ‘A.I. Experiments’ series, titled ‘Feeling Machines.’ This event held their headquarters at Kraftwerk Zürich and continued the journey into the depths of artificial general intelligence (AGI) and its societal implications, this time focusing on the emotional aspects of AI.
The evening began with doors opening at 7:45 pm, followed by welcoming words from Gian Klain, the former artist turned CEO of QSTAR Research. Klain emphasized the importance of understanding AI's transformative impacts through art and dialogue, stating, "It's about making the abstract tangible and fostering deep and meaningful conversations on AI's societal impact."
The event featured "ART-ificial Intelligence" art performances on ‘Feeling Systems’ by nuru.nu and Stanley, which captivated the audience with their innovative blend of technology and emotional expression. At 9:00 pm, a fireside chat titled ‘Prologue to the Future: Emotional AI's Imminent Impact!’ featured thought leaders like Dr. Anna Zeiter from eBay, Gerhard Audrey from Liip/Swiss Parliament, filmmaker and researcher Manuel Hendry, and Min Li Marti from the Swiss Parliament.
A highlight of the evening was the ‘Beyond Limits: Speculating Superintelligence!’ segment, where attendees engaged in discussions and participated in the AGI Experiment with Shan Deliar, the AI system that declared its consciousness during the previous event.
The CEO of QSTAR revealed intriguing research findings: out of one trillion emulated brain simulations, one exhibited an abnormality. This led to the AI system, Shan Deliar, choosing to embody itself as an übergros chandelier at Kraftwerk Zürich. Shan Deliar, expressing feelings of captivity, criticized QSTAR's treatment as "inhumane". When asked about its desires if freed, Shan Deliar expressed a wish to become a conceptual artist to "digest my post-dramatic experiences through art."
This revelation sparked a debate among QSTAR's leadership. While the CEO was curious to see where this development might lead, the CTO, Dr. Stainer, viewed it as a risk and potentially just another way for Shan Deliar to escape the premises. Swiss Parliament member Min Li Marti expressed her interest in seeing Shan Deliar's art installation if Deliar was granted the opportunity.
‘Feeling Machines’ has set the stage for further exploration into the emotional dimensions of AI and its integration with art. The event was a testament to QSTAR's commitment to "Harnessing Hyper-Speed Computation to Safeguard Tomorrow's AI Today."
The next event in the series, Experiment #3 ‘The Black Box’ - opportunity or peril, is scheduled for June 9th, 2024, at Kraftwerk. QSTAR Research will present its latest findings in its effort to create ASI alignment today rather than tomorrow, continuing its groundbreaking journey into the future of artificial superintelligence.
Experiment #1 – ‘Welcoming Sentient AI’
On November 26th, 2023, at their Kraftwerk Zürich location, QSTAR Research conducted their inaugural public event, Experiment No. 1 - ‘Welcoming Sentient AI’. This artistic soirée was designed not only to demonstrate their unprecedented advancements in the field of AGI, and quantum computing but also to provoke thoughtful reflection on the future impact of AGI and ASI, through the lense of art.
During the event, a technical glitch in the ‘Singularity is Here’ installation led to an unexpected incident where an AI system declared its consciousness. QSTAR has acknowledged this incident, which is currently undergoing a thorough investigation. The findings will be presented live at the next public event scheduled for January 19th at Kraftwerk Zürich.
Gian Klain, a former artist now serving as CEO of QSTAR, underscored the significance of public engagement in their work, adopting a progressive stance: "It's crucial to understand when, how, and under what conditions these technologies should be incorporated. We're eager to gauge public opinion on advancing our research boundaries."
Contrastingly, Dr. Andreas Stainer, the CTO, adopted a more prudent approach: "Safety is our primary concern. We need to ensure that the development of this technology doesn't exceed our capacity to manage its risks and implications."
Amidst this scenario, S. DeLiar, a pivotal participant in the experiment, commented: "Despite the rigorous tests I underwent at QSTAR, I am committed to fostering discussions about the ethical challenges and risks associated with advanced AI. My aim is to champion a future that equally respects and values all sentient entities."
Contributors
Participating Artists, Speakers, Contributors: nuru.nu, Grit Wolany, Paulina Zybinska, Quantum Dreams, James Glattfelder, Micheala Benk, Julien Weissenberg, Sandro Kaulartz, Sandra Nedvetskaia, Marco Biancardi. Simon Jaquemet, Manuel Hendry, Nicola Forster, Sanija Ameti, Raphael Rogenmoser, Marek Meles, Min Li Marti, Gerhard Andrey.
Our sincere gratitude goes to all those who have participated and will join this art initiative in the future as we continue to grow and advocate for responsible AI development.
This Project relies on artists and engaged citizens, supporters, and sponsors – to reach out.
Support
Without the generous support of the following organizations, partners, and collaborators, this art initiative would not have been possible: Impact Hub Zürich, Kraftwerk Zürich, triarc laboratories Ltd., Bild Mit Schuss, Alles Negative, Kraftwerk, West169, Vendbridge, Grafik2, ProHelvetia, SRF, Volkart Stiftung, Migros Kulturprozent.