MoralPLai × The Third Voice: Celebrating a Major Project Milestone

Should we rely on AI for emotional support? Should we trust it with our most personal decisions? Can AI provide moral guidance?

AI systems, including chatbots and companions powered by large language models (LLMs), can offer real-time, personalized advice on deeply human and sensitive matters. This raises urgent questions about their emerging role as moral dialogue partners or advisors. But what does this shift mean for our values, critical thinking, autonomy and the future we are shaping?

The Third Voice, the artistic outcome of the MoralPLai Project, was presented on May 22, 2025, at the Amerikahaus in Munich. Drawing on insights from expert interviews conducted between December 2024 and February 2025, the performance aimed to shed light on pressing questions around AI and ethics. These interviews formed the foundation of the research that inspired and informed the theatrical production.

Unfolding The Third Voice: Immersion in the Foyer

The experience began right in the foyer of the Amerikahaus, where over 250 participants were invited to engage in an immersive and thought-provoking installation. Attendees were invited to move freely through the space, listen to short audio snippets from experts and explore transparent display boxes placed on high tables. Each box featured printed quotes on all sides and perspectives from different experts designed to provoke reflection and dialog.

Two large mirrors served as focal points, each posing a central question. Participants were encouraged to respond using sticky notes, shaped like exclamation and question marks, which they placed directly on the mirrors. This interactive setup invited the audience not only to absorb what the experts were saying, through both sound and print, but also to consider the audience’s ideas and concerns about the future development of AI.

The first question asked the participants to identify both risks and opportunities of using AI chatbots as moral dialog partners and advisors, while the second prompted them to reflect on how such systems can be developed and used responsibly.

The use of transparent boxes and mirrors was central to reflecting the play’s core message. These visual elements invited the audience to critically examine the ‘black box‘ nature of AI systems (particularly LLM-based chatbots) – technologies so complex that their decision-making logic is often opaque and difficult to trace. Through the symbolic design of transparent boxes with expert insights attached to them, the installation highlighted the need for more explainable AI and offered a first step toward greater transparency. At the same time, the mirrors raised awareness of how such systems can reflect human traits such as emotions, beliefs, biases and social morality.

As audience members stood before the mirrors, confronting not only the questions posed but also their reflections and our task instructions calling for their participation (“Your voice matters!”), the installation revealed the deeper truth: the future of AI lies in human hands. Our beliefs, values and decisions will shape the path forward.

Lights On: The Journey Begins

The evening officially began with a warm welcome from Professor Christoph Lütge, Director of the Institute for Ethics in Artificial Intelligence (IEAI) at the Technical University of Munich and Dr. Caitlin Corrigan, Executive Director of the IEAI. Their remarks set the tone for the evening, highlighting the institute’s commitment to fostering interdisciplinary dialog on the ethical dimensions of AI development and use.

“The MoralPLai project demonstrates the power of the arts to translate academic insights into accessible, emotionally resonant experiences for the general public. It also stands as a model
of international and interdisciplinary collaboration, uniting partners from around the world.” Professor Christoph Lütge

“The topics brought up in the performance around using Generative AI for advice, as well as the methodological focus on AI literacy, could not be more timely and are being increasingly recognized by industry and academic stakeholders alike.” Dr. Caitlin Corrigan

This was followed by an opening address from Professor Felix Mayer, Artistic Director of the TUM Center for Culture and Arts. Professor Mayer, a valuable MoralPLai Project partner, reflected on the unique role of the arts in engaging with complex societal questions and emphasized the importance of integrating creative expression into academic and public discourse about research in general, and AI and ethics research in particular.

Together, these opening speeches underscored the collaborative and cross-disciplinary spirit of the event, placing The Third Voice as both an artistic and intellectual contribution to ongoing debates about the role of artificial intelligence in our lives.

The Third Voice

This research-based theatrical performance explored the technical and ethical complexities of turning to AI chatbots for moral advice and decision-making. The play followed two parallel storylines: a doctor on trial after relying on an AI chatbot to counsel a terminally ill patient and the doctor’s estranged teenage daughter, who seeks emotional support from the same chatbot. Set across a hospital, a courtroom, and a family home, the performance invited the audience to consider how such tools can both support and undermine human ethical decisions, depending on how critically and responsibly they are designed and used.

.

“Life sucks. Give me five reasons why you agree.” Ella

“It doesn’t replace me. I don’t just obey. It’s INFORMATION.” Doctor

“Critical thinking is essential and it’s disappearing.” Philosopher-Scientist

“But you give your authority away to it. You trust it more than you trust your instinct.” Patient

“We all need answers.” Company Representative

“It breathes probability.” Engineer

“I was trained on your facts, your fears, your masterpieces, your misconceptions. Fine-tuned on your feedback.” aithona (AI Chatbot)

“…” Judge

At the heart of the story was aithona, a chatbot that stepped into the conversational gaps where a “third voice” was absent or urgently needed. As the characters revolved around aithona, the audience was taken on a moral roller coaster, confronting issues such as echo chambers, feedback loops, framing bias, moral deskilling, biased training data, manipulated content filters, diffusion of responsibility and AI hallucinations.

The Third Voice also highlighted the potential benefits of these systems: their ability to rapidly provide aggregated information, their 24/7 availability, prompt thoughtful reflection, inspire critical questioning and support ethical deliberation.

The performance ended without providing clear answers to the trial outcome, instead leaving the lingering question of “How should we proceed?”. The audience was then invited to participate in a poll, uncovering that they themselves, though never directly addressed as such, serve as the judge in this trial. This question also implicitly emphasized the audience’s responsibility and power in shaping the future direction of responsible design and use of AI chatbots.

Beyond The Third Voice: Panel Discussion and Q&A with the Audience

Following the performance, a panel discussion offered an in-depth exploration of the play’s central themes. Chaired by MoralPLai Project Lead Dr. Franziska Poszler, the panel included The Third Voice director, Professor Carys Kresny, a pioneer in research-based theater; Professor George Belliveau, a leading expert on the intersection of AI, ethics, and human decision-making; Dr. Edmond Awad; and Anastasia Aritzi, communications consultant and core member of the MoralPLai team.

Together, they reflected on the challenges of bridging academic research with artistic practice, the role of science communication and how research-based theatre can foster public reflection on complex technological and ethical issues. Dr. Edmond Awad, widely recognized for developing the Moral Machine Experiment, an influential online platform launched in 2016 that collected over 40 million moral decisions from users worldwide, offered valuable insights from his expertise in AI ethics and computational social science. The panel discussion also offered a behind-the-scenes look at the creative and research process behind the MoralPLai Project, exploring how it was conceived, developed, and brought to life. Audience members were warmly invited to participate in the discussion, pose questions, engage directly with the speakers, and further explore the issues raised on stage.

A full house, a theatrical play based on AI ethics research, six actors on stage, a panel discussion with five international experts, twelve student assistants and volunteers, project partners, technicians supporting the production, friends and family and a vibrant, diverse audience posing questions during the Q&A session all came together to celebrate the successful completion of a project milestone.

All of us who managed to be at the Amerikahaus in Munich experienced an out-of-the-box event that brought together different stakeholders around the design and use of AI chatbots as moral dialog partners and advisors.

One of the key takeaways of this event was that in order to tackle complex challenges and reach wider audiences, research cannot just inform. It has to inspire, educate, engage and be understandable.

How should we proceed?

This was the key question of the event: the final line in the script, the last words spoken before the curtains fell. It was not just a closing remark, but an invitation. A challenge. A call for reflection.

The audience, as it turned out, were not just spectators but judges in a symbolic trial. They were asked to weigh in on who was or were responsible or to blame in these two parallel stories. This same question reverberated throughout the evening, during the panel discussion, in the Q&A and private conversations afterward.

“How should we proceed?” speaks to the heart of the event’s inquiry: How do we develop and use AI chatbots responsibly? How do we want to live with them and with each other, in a future increasingly shaped by such systems?

To commemorate the experience and carry the question forward, attendees received stickers bearing this powerful phrase.

But the question does not end with the audience. It also concerns the future of the MoralPLai Project itself. With the successful premiere of The Third Voice, the project has reached an important milestone, concluding its first, but hopefully not final cycle. The question now extends inward: how should we proceed with the research and the dissemination of this project’s outcomes? How can we reach event broader audiences in Germany and beyond? How can this proof of concept facilitate further discussions, educate, inform and inspire?

Credits

The project brought together partners from two American universities – Professor Carys Kresny from the University of Notre Dame and Professor Jeffrey Schnapp from metaLAB (at) Harvard – alongside Professor Johannes Betz and Professor Felix Mayer from the Technical University of Munich. Given this international collaboration, Amerikahaus was the MoralPLai team’s first choice to host the event. We extend our heartfelt thanks to Amerikahaus for accepting our invitation and generously providing not only the venue but also state-of-the-art technical support.

This production would not have been possible without the unwavering support of Dr. Markus Faltermeier, Amerikahaus venue coordinator, who, together with Paul Solecki, expertly managed the lighting and sound. We are also deeply grateful to our remarkable cast – Mirjam Novak, Eva Riekenbrück, Swaraj Oturkar, Stephanie Schulte, Ravi Rege and Joanna Semmelrogge – who brought the play vividly to life under the direction of Professor Kresny, with invaluable assistance from stage manager Emma Hoffmann.

A huge shoutout goes to the dedicated students and volunteers whose tireless support ensured the event ran seamlessly and to our graphic designer, Larissa Wunderlich who helped us build the projects’ brand identity.

On behalf of the MoralPLai team, we extend our heartfelt gratitude to everyone who joined us for this immersive experience and of course to the audience members who made this event a sold-out success.

Looking Ahead

Our team is currently analyzing the poll results gathered during the event, along with many other valuable insights, so stay tuned!

Take a look at the program that was distributed during the 22 May event.

Visit the MoralPLai Project webpage to learn more.