Generative AI as a Tool for Design Ethics Education

In an article published in the journal Cybernetics and Human Knowing, Thomas Fischer describes an innovative assignment for his SUSTech School of Design undergraduate course DS208 Design Ethics. Rather than focusing on cautionary tales or abstract ethical dilemmas, Tom invites students to envision desirable futures using AI-generated imagery in the solarpunk aesthetic—an optimistic, eco-futurist movement centered on sustainability and social justice.

Tom’s approach deliberately avoids what he calls “doom-mongering.” Today’s students face what he terms a “dystopian polycrisis” such as climate change, economic instability, youth unemployment etc., challenges today’s design students did not create but must confront. Rather than adding to feelings of hopelessness, the course aims to foster agency, creativity, and resilience.

The assignment draws on Herbert Brün‘s “Desires Exercise,” asking students to imagine a desired future state and then reason backwards: what transformations would have been necessary to bring it about? This retroductive approach shifts engagement from critique toward constructive speculation.

Students are assigned ethics themes ranging from ocean health to automation, from gaming to globalization. After researching and presenting their topics, they create four-panel visual narratives depicting hopeful futures using AI-generated imagery. The solarpunk aesthetic—introduced to the course through the animated short Dear Alice—provides a visual framework emphasizing harmony between technology and nature.

Student works have explored environmental simulation games teaching ecological responsibility, decentralized energy networks empowering rural communities, and post-war cities rebuilt around peace education. Student reflections reveal genuine shifts in perspective, with one noting: “Design is never neutral—especially when entangled with systems of violence, war, and power.”

In a recursive turn characteristic of cybernetics, the assignment surfaces ethical dilemmas inherent in its own methodology. AI image generation carries significant environmental costs, concentrates wealth and power, potentially devalues creative labor, lacks transparency, and raises intellectual property concerns—particularly when prompts reference specific artistic styles.

Tom makes these tensions explicit rather than hiding them. The question of whether pedagogical benefits justify these costs becomes itself a subject of design ethics inquiry. As one student critically observed, AI art relies on training data from artists who receive neither acknowledgment nor compensation. He offers no definitive answer about whether educators should adopt such tools. Following Heinz von Foerster‘s insight that truly ethical questions are, in principle, undecidable, he argues each educator must make this determination personally.

His provisional resolution is to continue the assignment while integrating its inherent dilemmas into course content. This approach frames design ethics not as a destination to reach but as an ongoing negotiation—simultaneously enabled and challenged by emerging technologies. Each generation of designers must confront these tensions anew, developing their own answers to questions that resist easy resolution.

Reference

Thomas Fischer (2026). Positive feed-forward: The use of generative AI for visual storytelling in design ethics education, Cybernetics and Human Knowing, 32(3-4): 111-132.