Public interest technology, technology for the good, digital public infrastructure – no matter how you call it: we need more of it. Therefore, it was a pleasure to welcome Audrey Tang to our Bertelsmann Stiftung office in Berlin to discuss plurality, technology, and democracy and to learn from Taiwan’s approach.

“More public interest technology” is becoming a prominent slogan. After years of semi-monopolistic Big Tech, path dependencies, lock-in effects, polarization, and division, the demand for a better tech ecosystem is immense. With reframe[Tech], we want to focus on strengthening and improving such digital public infrastructure. A good starting point for such an endeavor is to look at what other like-minded states and actors are doing and learn from their experiences. That is why it was a pleasure to interview Audrey Tang, former Digital Minister of Taiwan, civic hacker, and author to explore how the concept of plurality can help create technologies that support democracies and are developed for and with the public interest in mind. The occasion was the book launch of “Plurality”, co-authored by Tang with Glen Weyl (and over 100 contributors) that provides a framework on how to use technology to strengthen democracy.

***The interview is also available as a video on YouTube. You can find the link below this post.***

The concept of Plurality: Collaborating across social differences

In “Plurality,” Tang and Weyl describe the concept as a third way beyond libertarianism and technocracy. According to Tang, plurality means collaborating across social differences. She shared an example from Taiwan: “When Uber X [a ridesharing company/app, Editor’s note] came to Taiwan, we had taxi drivers and Uber drivers fighting each other. Social media at the time fueled polarization instead of building bridges between the different camps.” Instead, we need technology that enables people to deliberate and hear diverse opinions, ideas and building consensus. Her analogy: digital infrastructure should serve as a public square, not as noisy nightclubs. How will we get there: “by building pro-social social media that floats consensus or at least builds bridges.”

In an interactive web publication we demonstrate how so called “bridging algorithms” can contribute to a better digital discourse  –  using the Uber example from Taiwan, among others.

Taiwan tried this via the open-source tool “Pol.is” as part of the “vTaiwan”-project. “vTaiwan” is a decentralized consultation process that combines online and offline interactions, allowing Taiwan’s citizens and government to discuss and deliberate on national issues together. The initiative enables nuanced discussions that lead to consensus on contentious issues, as the conflict around the ride sharing app Uber. “When people deliberate and hear diverse ideas, they become inoculated against polarization,” Tang explained. This approach, she believes, can serve as a public digital infrastructure that defends against polarization and fosters co-creation.

The role of civil society: trusted, credible, neutral information processors

Tang emphasized the pivotal role of civil society in driving this shift. “Foundations and the not-for-profit sector are essential because they are trusted, credible, neutral information processors,” she said. In Taiwan “Cofacts”, a collaborative fact-checking tool allows anyone to report scams, phishing, and disinformation to an open-source dashboard. This tool is trusted by companies and the public alike because it is operated by civil society organizations, independent from the government and profit interests.

Tang argued that civil society should proactively demonstrate better solutions rather than just protesting against (tech) issues. “In the sunflower movement [The Sunflower Movement was a 2014 Taiwanese student protest against a trade deal with China, Editor’s note], we were not protesters, we were demonstrators. Protesting is being against something, demonstrating is showing something better,” she stated. She acknowledges that the state must provide (financial) support for civil society so that it is able to build alternative tools in order to effectively show examples of how different a public interest technology could be.

Decreasing Big Tech’s power concentration through “public money, public code”

Tang believes that public code investments are vital to countering Big Tech’s power. “Public money should fund public code” she explained. In Taiwan, this principle is implemented through various initiatives, ensuring that public sector technology remains open and interoperable. For instance, the country insists on Open Document Format (ODF) to ensure interoperability. “We didn’t force everyone to use LibreOffice or OpenOffice, but we required that Google Docs and other services support ODF for import and export,” Tang said. This approach ensures that proprietary vendors cannot lock in and exclude future vendors, increasing the public sector’s negotiation power. Tang advocates for more foundations and NGOs to complete this picture, “so that we get technology that fits to the societal needs, not asking society to wait for technology to change.”

For additional perspectives on this issue, see our related panel discussion titled “Scaling Collaborative Technology to Enable the Future of Multilateralism”, which explored how collaborative technology can be leveraged to advance multilateral cooperation and address global challenges.

Taiwans approach to generative AI

And because generative AI is omnipresent, I couldn’t pass up the chance to hear Tang’s insights on how foundation models can embody these discussed values of openness, diversity and public interest. Taiwan has developed its own foundational model, the “Trustworthy AI Dialogue Engine” (TAIDE). “We align TIADE by asking people how they want AI to interact, creating a constitutional document based on a Wiki survey,” Tang explained. A constitutional document for LLMs is a collaboratively created guideline that ensures these models operate in alignment with a society’s values and ethical standards. This process, known as “Alignment Assembly,” ensures the AI model aligns with Taiwanese norms and values – setting an example for how public interest technology can be developed in a way that reflects societal values and promotes inclusivity and transparency.

The interview with Audrey Tang and her perspectives emphasize the critical need to strengthen responsible digital public infrastructure. On the path to a tech ecosystem oriented toward the common good, we can learn a lot from Taiwan: building bridges, strengthening civil society, demonstrating better solutions, and ensuring interoperability are essential steps toward a more inclusive and democratic digital future.

Die Kurzzusammenfassungen der Artikel sind unter einer Creative Commons Namensnennung 4.0 International Lizenz.