Our Research

Our first test of generative polling was designed to get a deeper understanding of how people think Using an AI-enabled generative polling tool we took 112 participants from the West Australian electorate of Curtin through a structured, text-based conversations about housing taxes like negative gearing and the capital gains tax discount. Participants answered open and closed questions in their own words and then evaluated one another’s responses, with the tool using machine learning to cluster, summarise, and quantify the ideas that resonate most strongly across the group.

The results showed us not only what people thought but gave insight into why people thought it. We’ve learnt some important lessons from this first test, and are using these learnings to build our research methods further.

36 participants from Toowoomba, Queensland took part in a full-day investigation into what divides and connects Australia’s generations. Through structured dialogue, expert input, data exploration, and creative exercises, participants stress-tested their assumptions and examined the realities of the “generational contract”, pinpointing both shared values and clear points of tension around housing, climate, tax, and opportunity.

The findings offer concrete pathways for narrowing generational divides: clearer communication across age groups, more opportunities for real-world intergenerational contact, and a stronger focus on the common ground that already exists. The session provides an early blueprint for how communities can hold rigorous, honest, and constructive conversations about complex issues.