What if AI literacy weren’t developed in research labs alone, but co-designed with the communities it is intended to serve?
That question brought together 54 participants from five countries for Commun-AI-ty: Building Science Communication Skills for Explaining AI in Science, a 3.5-day workshop held at the University of Michigan from June 23–26, 2026. Made possible by a more than $93,000 award from Schmidt Sciences and hosted by the Michigan Institute for Data and AI in Society (MIDAS), the workshop challenged researchers, educators, and designers to rethink how artificial intelligence can be communicated through a human-centered, participatory design process.
Throughout the workshop, participants with expertise in AI, education, design, and science communication worked in interdisciplinary teams to develop engaging ways to explain AI in science. Rather than designing educational tools in isolation, they collaborated directly with five middle and high school teachers, whose perspectives and feedback guided the development of each project through an iterative user-centered design process.
The workshop culminated in the creation of seven interactive web applications designed to make AI concepts more accessible to learners. The prototypes explored topics including effective prompt writing, how large language models (LLMs) generate responses, evaluating the reliability and limitations of AI systems, AI-assisted learning, digital twins, multi-agent systems, and AI-powered scientific discovery. The projects—Class Plus, How AI Are You?, Multi-Agent System Demystifier, Newton, Professor Frankfurter, Prompt Minion, and Teach AI How to Think—are available through the Commun-AI-ty showcase website: commun-ai-ty.dept.ic.ac.uk and provide engaging resources for educators, students, and the broader public.
“The most valuable moments came from listening,” said workshop organizer Long-Jing Hsu, Schmidt AI in Science Fellow at the University of Michigan. “The teachers continually reminded us that effective science communication doesn’t begin with the technology—it begins with understanding the people we’re communicating with.“