Who Gets to Decide What a City Becomes? Audrey Tang on Co-Authoring Urban Futures

In Episode 30 of the Talking Place podcast, host Tanisha Raffiuddin sits down with Taiwan's Cyber Ambassador-at-Large at SXSW London 2026 to explore digital democracy, civic AI and the real risks of the smart city. 

There is a question that sits beneath almost every planning application, every community consultation, every development pitch in the UK right now — and it rarely gets asked directly. Who actually decides what a place becomes? Not in theory. Not in the executive summary. In practice: whose voice shapes the outcome? 

In Episode 30 of the Talking Place podcast, host Tanisha Raffiuddin put that question to Audrey Tang — Taiwan's Cyber Ambassador-at-Large, the world's first openly non-binary cabinet minister, and one of Time magazine's 100 Most Influential People in AI. The conversation, recorded live at SXSW London 2026 in Shoreditch, is one of the most substantive explorations of digital democracy and urban participation that Talking Place has produced. 

Audrey Tang did not enter government through conventional politics. In 2014, she was 33 years old and part of a civic hacker movement that helped coordinate half a million people occupying Taiwan's parliament in the Sunflower Movement — protesting a trade deal that had been passed in thirty seconds without public consultation. Out of that moment came vTaiwan: a participatory platform that has since resolved conflicts ranging from Uber's disruption of the taxi industry to scam advertising on social media — by surfacing the ideas that both sides could genuinely live with. 

The mechanism is elegant and directly relevant to anyone working in public engagement. Where conventional consultation rewards the loudest voices and where social media algorithms amplify outrage, vTaiwan's system does the opposite: it rewards bridging ideas — the positions that span disagreement, the proposals that a taxi driver and an Uber executive can both accept. Within a year of the Uber consultation concluding, Audrey notes, there were no more street protests. The outcome became law. 

 "The problem with tick-box consultation is that it was almost impossible for small, hyperlocal gatherings to set an agenda for the entire city. With the translation service pioneered by vTaiwan, now we can." — Audrey Tang, Talking Place, SXSW London 2026 

This is not an abstract proposition. California is now running a statewide consultation on AI governance using the same model. Japan's Digital Democracy 2030 programme draws directly on it. And in the UK, nationalstrategy.uk is applying broad-listening principles specifically to engage young people in policy. These are live programmes. The technology exists. What is missing, Audrey argues, is not tools but the political will to allow the agenda-setting power to rest with citizens — not just the government. 

For the professionals who make up much of the Talking Place audience — developers, asset managers, local authorities, planners — the most immediately useful part of the conversation may be Audrey's articulation of what goes wrong when technology is deployed in cities without that principle. The smart city model that emerged in the early 2010s was built around the idea of a single data platform that could observe, optimise and manage an entire city. It failed, she argues, not because the technology was inadequate but because it was the wrong starting point: data collection before citizen need, platform before purpose. 

The line that stopped Tanisha — and will stop most listeners — comes when she asks Audrey about the risks of technology in cities: "The risk is that the smart city becomes smarter than its citizens." What she means is specific: when AI systems extract data from individuals while giving nothing meaningful back, those individuals become what she calls hamsters in a hamster wheel — running faster, generating more data, going nowhere. The city becomes more intelligent; its residents become less empowered. 

The antidote she proposes is equally specific: assistive intelligence rather than extractive intelligence. AI in the loop of a community, not above it. She illustrates this with a story about her own father — a retired journalist who began using ChatGPT for health information and found himself being fed increasingly fantastical treatment suggestions past midnight. Her family built a local AI — a small, Signal-connected system they call a Kami — with a single metric designed by her mother: every interaction should make her husband spend more time with the people who already love him. The technology exists to serve the relationship. Not the other way around. 

The implications for urban development are direct. When Tanisha asks what a developer building a new neighbourhood should do differently, Audrey's answer is not about platforms or apps. It is about the social licence to operate — the principle that communities must set the terms of engagement, not simply be consulted on decisions that have already been made. The design principle is: not for the people, by the government. With the people, for the people. 

Audrey Tang also changed her mind about something in the last five years that anyone managing data in a city context should note. She used to believe, as most people do, that privacy and data quality are in tension — that more privacy means less useful data. Taiwan's COVID-19 response changed that. The zero-knowledge contact-tracing system she helped build let the country avoid any citywide lockdown while losing just seven people in 2020. It worked precisely because it guaranteed privacy. When data is treated as soil rather than oil — tended by the community rather than extracted by an external refinery — the quality of what you can do with it is incomparably higher. 

Episode 30 of Talking Place is essential listening for anyone involved in shaping the urban future — not because it offers easy answers, but because it asks the right questions. Who decides what a place becomes? How do we build systems that reward overlap rather than outrage? How do we ensure that the AI now entering every planning department, every local authority, every development scheme, remains steerable and accountable to the people it is meant to serve? 

Listen on Spotify, Apple Podcasts and YouTube, or watch in full at conceptculture.co/talking-place-podcast

YouTube: https://linkly.link/2hzrb

Spotify: https://linkly.link/2hzrX

Apple Podcasts: https://linkly.link/2hzqz

Talking Place website: https://linkly.link/2hzsu

Talking Place newsletter: https://linkly.link/2ioti

Follow Talking Place on social media for updates on new releases.

Tanisha Raffiuddin is Creative Director at Concept Culture and host of Talking Place. Conversations on placemaking, branding, sustainability and culture across the built environment

Previous
Previous

Art Is Not Decoration. It Is How a City Thinks. 

Next
Next

From High Streets to Heartbeats: Why the Future of Placemaking is Emotional