New research: A Framework for Welfare Resilience in the Age of AI

Read more

New research: A Framework for Welfare Resilience in the Age of AI

Why Scenario Planning?

The future of AI and economics isn't a single predetermined path. It is a landscape of possibilities, and one of the most uncertain periods in recent history.

We treat AI-driven economic disruption the way we approach pandemic preparedness: the most consequential scenarios aren't always the most probable ones. If we only plan for conservative, incremental change, we risk being catastrophically underprepared for more transformative outcomes.

The future of AI and economics isn't a single predetermined path. It is a landscape of possibilities, and one of the most uncertain periods in recent history.

Scenario planning helps us:

Expand our collective imagination beyond business-as-usual projections

Stress-test policies against a range of futures, not just one forecast

Identify where our institutions might break and what needs reinforcing now

Build societal resilience by preparing for economic shocks before they arrive

Surface the questions, pressure points, and choices that will matter most

Build common vocabulary across economists, technologists, policymakers, and other communities

Our Workshops

Our workshops fall into two types. Some establish intellectual foundations — bringing together economists, philosophers and social scientists to stress-test the biggest questions about AI and economic futures. Others put that thinking into practice, working with specific audiences — national policymakers, city leaders, central bankers, private sector institutions, governments across the Global South — on the concrete decisions they face. What connects them is the methodology: the same rigorous scenario approach, adapted to each room.

Foundations

The Economics of TAI

San Francisco, September 17th 2025

Co-convened with Anton Korinek, Erik Brynjolfsson, and Ajay Agrawal, this workshop brought together 40+ leading economists and AI lab representatives to stress-test the macroeconomic implications of transformative AI. Participants worked through four high-growth scenarios, exploring how different speeds and shapes of AI adoption could reshape labor markets, productivity, and the broader economy.

Dignity of Work in the Age of TAI

San Francisco, March 17th 2026

Co-convened with Notre Dame, this workshop brought AI lab participants together with voices from faith and philosophy to confront a question economics alone can't settle: what happens to human dignity when machines can do all the work? Participants compared two sharply different futures — one built around guaranteed income, the other around guaranteed jobs — to surface the deeper tensions beneath the policy debate.

Global Economic and Financial Implications of AI

Washington, DC, December 10–11, 2025

Co-convened with the IMF and Anton Korinek's Economics of Transformative AI Initiative at the University of Virginia, this workshop brought together roughly 50 IMF staff, central bankers, and outside AI and economics experts to stress-test how a genuinely transformative AI transition—not just another technology shock—would ripple through growth, labor markets, and financial stability.

Participants worked through two diffusion scenarios that shared the same aggressive assumption about frontier AI capability but diverged on how fast societies absorb it: a baseline world where regulation, infrastructure bottlenecks, and public pushback slow adoption to a crawl, against a runaway world where automation sweeps across sectors with almost no friction.

Groups were tasked with mapping the economic fallout of each scenario under current policy, then asked how governments should respond—surfacing where fiscal, monetary, and financial-stability tools would strain or break under each path.

In Practice

Korea's AI Windfall: Avoiding the K-Shaped Economy

Seoul, July 2026

Economists, technologists, and policymakers gathered in Seoul at a workshop co-organized with the Bank of Korea to test whether Korea's next transformation can hold its social contract together.

When America's Economy Meets AI

Washington, DC, June 11, 2026

Co-convened with the Brookings Center on Regulation and Markets, the Foundation for American Innovation, and the Peterson Institute, this workshop placed economists, AI researchers, and policymakers in a 2030 America where AI diffuses rapidly through white-collar work. GDP growth is strong, equity markets are soaring—and the middle class is pushed into precarity. The exercise took transformative AI's arrival as a given and tested something narrower and more practical: if this is where current trajectories lead, has the country done anything to prepare?

Preparing Switzerland for Transformative AI

Zurich, May 7, 2026

Co-convened with Elliott Ash of ETH's Law, Economics and Data Science faculty and the Albert Einstein School of Public Policy, this workshop placed economists, technologists, social partners, and policymakers inside a 2030 Switzerland where AI has automated much of the high-skilled professional work the country's competitive advantage rests on. Advising a federal council with a limited window, participants converged on an uncomfortable finding: the bottleneck is institutional, not intellectual—the policy levers exist, but the consensus-driven institutions that would have to pull them are built for gradual change, not an exponential one.

AI Economic Preparedness In Cities

Madrid, April 2026

Co-convened with Bloomberg CityLab, this workshop brought together city leaders, urban economists and policymakers to stress-test what AI-driven economic disruption looks like when it lands in a specific city. Participants worked through the scenario where a mid-sized European city faces the sudden automation of 20,000 jobs, confronting the immediate trade-offs between stabilizing local labor markets, protecting municipal finances, and maintaining public trust with uncertain national support.

Policy Strategies for TAI

London, February 5th 2026

Built for UK policymakers, this workshop dropped participants into a bespoke scenario set in October 2030 — where AI-driven displacement of high earners has eroded the tax base even as GDP keeps climbing. Eight groups were tasked with the same brief: prepare the Prime Minister's three-point plan to strengthen the social contract.

Our Approach : Ideal Types and Worldbuilding

Our scenarios are not forecasts. They are ideal types, in the Weberian sense: deliberately simplified worlds that exaggerate certain dynamics in order to make underlying forces visible.

In practice, this means:

Our scenarios are not forecasts. They are ideal types, in the Weberian sense: deliberately simplified worlds that exaggerate certain dynamics in order to make underlying forces visible.

In practice, this means:

Pinning different futures next to one another

Treating assumptions as scaffolding, not commitments

The key is contrast. The more differences between the scenarios, the better. Similar-sounding scenarios don't help us explore the possibility space or stress-test our thinking.

It is much easier to challenge a concrete world than to reason from a blank page. The narratives and worldbuilding elements are there to be tested, questioned, and designed to be helpfully wrong in their concreteness.

Importantly, these scenarios are not mutually exclusive. Societies may begin on one trajectory and later slide into another. Part of this work is developing indicators that help us reason about which future we may be moving toward, and how quickly.

The key is contrast. The more differences between the scenarios, the better. Similar-sounding scenarios don't help us explore the possibility space or stress-test our thinking.

Positive and normative exploration

Depending on the audience, scenarios can balance positive economics and normative economics in different ways.

Positive Economics

What will happen?

These scenarios focus on describing likely effects and outcomes given certain conditions, helping us understand economic dynamics in the absence of intervention. They are useful in identifying vulnerabilities and where our systems strain.

Normative Economics

What should we do?

These scenarios center on interventions, policies, and choices, exploring what actions might be desirable or necessary given certain conditions. They are useful in identifying foreground policy design, values, and trade-offs.

Economists, policymakers, and civil society actors often benefit from different mixes of these perspectives. Our workshops are designed to adjust that balance deliberately.

Beyond Equilibrium: The Transition Challenge

Ideal types are useful for clarifying end states and identifying broad patterns. They help build intuition about how different economic systems might function. However, an important challenge remains: explaining how we move from today’s economy to those possible futures. Understanding these transition paths will be a key focus of future work at Windfall.

Charting a robust course between these possible futures requires grappling with additional questions, including:

How fast will progress actually unfold?

How much time will our institutions have to respond?

What's the right sequence for policy interventions?

What feedback loops and cascading effects might we encounter along the way?

The scenarios help us see the destinations. The next phase of our work focuses on mapping the transition paths, feedback loops and points at which early decisions close off later options. We are bringing narrative and historical expertise alongside economic analysis to make those paths as concrete as the scenarios themselves.

Preparedness is not about predicting the future. It is about being ready for more than one.

What’s Next for Windfall

Windfall’s next phase of work will focus on convening and listening. We will be running a series of workshops to engage people across different geographies in thinking through the economic implications of transformative AI. These workshops are designed to build shared understanding and create bridges between technologists, economists, policymakers, civil society, and other communities that are often siloed from one another.

Our upcoming workshops will begin in New York and Washington, DC, with plans to expand to additional regions. We are actively exploring workshops in the Global South and Asia, where perspectives on AI, economic disruption, and redistribution are often underrepresented in global conversations.

We are also collaborating with the Partnership on AI on a joint workshop this summer, followed by a public report later in the year. In this collaboration, Partnership on AI will lead work on drivers and recommended actions, while Windfall focuses on scenario development and facilitation.

To explore partnering or supporting this work, get in touch at contact@windfalltrust.org

Getting ahead of AI's economic disruption