Gideon Lichfield, a writer for Substack’s Futurepolis and former Wired editor-in-chief, dives into the tense relationship between AI and democracy. He discusses how AI models, controlled by a select few, impact governance and the economy. Insightfully, he explores solutions like participatory budgeting and technology’s role in enhancing civic engagement. Lichfield shares experiences with risk gaming scenarios involving deepfakes, and emphasizes the importance of creating public-interest AI that prioritizes societal good over profit.
The centralization of AI technology by a few large companies raises significant concerns about accountability and public representation.
Gideon Lichfield's insights from risk gaming events reveal the complexities and chaos policymakers face in managing misinformation and competing demands.
Promoting public interest AI can ensure technological advancements serve the common good, although challenges in funding and regulation remain prevalent.
Deep dives
The Centralization of AI and Its Implications
AI is reshaping various aspects of society, but its centralization into the hands of a few large companies raises critical questions about control and accountability. A few dominant model makers benefit from this technology, leading to concerns that public interests are overlooked. The podcast discusses whether public involvement in AI development should be increased to ensure fair representation. This tension highlights the need for a balanced approach where technological advancements benefit a broader audience.
Experiences in Risk Gaming: Chaos and Complexity
Through Gideon Litchfield's experiences with the risk gaming events, insights into managing chaos and complex decision-making in critical scenarios were illustrated. The game 'Deep Faked and Deep Six' simulated the confusion policymakers face during election subversion attempts, demonstrating how misinformation complicates real-world decision-making. Similarly, the 'Powering Up' game reflected the difficulties city mayors encounter when trying to balance demands from car manufacturers and central government pressures. These scenarios emphasized the chaotic environment of governance and the challenge of making informed choices amid competing agendas.
Participatory Democracy: Challenges and Opportunities
Gideon discusses the importance of participatory democracy and civic engagement in modern governance, highlighting the successes and limitations of initiatives like participatory budgeting. For instance, Seattle's budgeting process aimed to give marginalized communities a voice but faced implementation challenges. Similarly, a mini citizens’ assembly in Tennessee reached consensus on gun safety measures but struggled to engage lawmakers for meaningful adoption. This illustrates the need for genuine buy-in from established political structures to make participatory processes impactful.
The Need for Public Interest AI
The concept of public interest AI proposes creating AI models and infrastructure that serve the public good rather than being solely profit-driven. By drawing parallels to public services like the NHS and the BBC, the discussion emphasizes the potential of public AI to ensure accessibility and promote democratic values in technology. However, questions arise about funding and the regulatory environment needed to support such initiatives, especially in more skeptical environments like the US. Ultimately, this approach aims to empower broader access to technology without sacrificing ethical standards.
Governance in the Age of Rapid Technological Change
The podcast underscores the disconnect between the speed of technological advancements and the sluggish pace of governmental processes in lawmaking. The existing political framework struggles to adapt quickly to changes brought by innovations such as AI, often resulting in outdated regulations. To remain effective, governments must update their procedures and foster an environment that accommodates the fast-paced nature of modern technology. This call for reform highlights the need for a governance model that balances established institutions with the agility required to navigate contemporary challenges.
AI and democracy are in great tension with each other. AI models are built by a priesthood of research specialists, unmoored from the will of the public. Yet, these very models are increasingly running important parts of the economy and increasingly government. How do we take advantage of these new capabilities without losing control of them?
That’s the debate at the center of our conversation today with Gideon Lichfield, the writer behind the Substack “Futurepolis” and the former editor-in-chief of Wired. Through his recent writings, Gideon has explored what a public option might look like with AI, how participatory democracy can be buttressed by new organizational and technical designs, and the tactical changes needed to make it much easier for government to procure software.
Joined by host Danny Crichton and Riskgaming director of programming Laurence Pevsner, we first talk about Gideon’s two recent experiences playing our scenarios on the Chinese electric vehicle market and AI deepfakes. Then we pivot to a broader conversation on the future of governance, discussing everything from participatory budgeting and liquid democracy to balancing between technocracy and democracy while remaining optimistic about the future.
