The future is here, and it's powered by AI. But as we embrace this powerful technology, a crucial question arises: how will it shape our economy? The answer is uncertain, leaving policymakers with a daunting task.
Since the launch of the Anthropic Economic Index, we've witnessed a significant shift. Users are increasingly delegating tasks to AI models like Claude, reducing the need for human collaboration. As AI systems become more independent and employers adopt them for productivity gains, this trend is set to accelerate. The implications for the workforce are unclear, leaving policymakers with a challenging dilemma.
But here's where it gets controversial... How should they respond? It's a complex issue with no easy answers. There's uncertainty about the scale of the transition and a wide range of opinions on how to manage it. Despite these challenges, formulating ideas for potential economic scenarios is imperative.
Over the past year, we've collaborated with economists and policy experts worldwide to advance this discussion. We engaged with thinkers from diverse political backgrounds to generate a broad spectrum of ideas.
Below, we explore nine categories of policy ideas, covering workforce development, permitting reform, fiscal policy, and social services. While we can't predict the optimal policies, we're committed to transparency and sharing ideas openly.
Matching Policies to Scenarios:
The nature of AI's economic impact will determine the policy responses needed globally. We've organized these initial ideas into three categories:
Policies for Nearly All Scenarios: These ideas are relevant regardless of AI's disruption level. They include upskilling workers and students for emerging jobs and reforming permitting processes to enhance productivity.
Policies for Moderate Scenarios: Here, AI leads to measurable wage declines and job losses. More substantial fiscal support for displaced workers might be necessary. Taxes on automation could offset negative externalities imposed on workers.
Policies for Faster-Moving Scenarios: These proposals are ambitious, designed for dramatic job losses and worsening inequality. Ideas include using sovereign wealth funds to give citizens stakes in AI revenues and generating new government revenue streams.
These proposals don't represent Anthropic's official policy positions, but we're excited by the diversity of ideas and hope they spark further research and debate.
Policies for Nearly All Scenarios:
Upskilling through Workforce Training Grants: Abigail Ball, Executive Director of American Compass, presented the Workforce Training Grant at our DC Symposium. This proposal directs public resources towards on-the-job training, with governments providing annual subsidies to employers creating formal trainee positions with structured programs.
Reform Tax Incentives for Worker Retention and Retraining: Tax policy can incentivize employers to retrain and retain employees. Revana Sharfuddin of the Mercatus Center argues that the US tax code favors physical capital over human capital investment. She proposes reforms to reduce the cost of retraining relative to layoffs, helping workers whose positions are at risk.
Close Corporate Tax Loopholes: Tax policy expert David Gamage has outlined reforms to prevent AI transformation from straining government budgets. These include closing the "partnership gap" and modernizing tax allocation to combat profit shifting from digital and intangibles-based business models.
Accelerate Permits and Approvals for AI Infrastructure: Anthropic advocates for reforming permitting processes in the US and allied nations. Accelerating these processes is crucial for developing the infrastructure to train and deploy frontier AI. Reforms will unlock investment, economic growth, and job creation. Failing to do so could slow productivity and introduce national security risks.
Policy Ideas for Moderate Scenarios:
Establish Trade Adjustment Assistance for AI Displacement: Economists are exploring how the Trade Adjustment Assistance model can be adapted for AI-driven labor market disruptions. Ioana Marinescu of the University of Pennsylvania views TAA-like "AI insurance" as a mechanism to support those losing jobs due to AI.
Implement Taxes on Compute or Token Generation: University of Virginia economists Lee Lockwood and Anton Korinek propose studying taxes on token generation, robots, and digital services. These taxes offer benefits and risks depending on AI's development stage. A tax on AI-generated tokens sold to end users might be desirable when humans remain dominant consumers.
Policy Ideas for Fast-Moving Scenarios:
Create National Sovereign Wealth Funds with Stakes in AI: Proposals aim to give citizens and governments greater stakes in AI's economic returns. Sovereign wealth funds could enable states to invest in AI-related assets, shaping the sector's behavior and distributing AI-derived wealth more equitably.
Adopt or Modernize Value-Added Taxes: Six out of the G7 countries and 37 out of 38 OECD countries have national value-added taxes (VATs). As AI transforms the economy, labor's share of value production might decline, making a shift towards taxing consumption necessary to fund core government activities.
Continuing the Conversation:
Earlier this fall, Anthropic announced a $10 million commitment to scale up the Economic Futures Program. This investment will support empirical research on AI's economic impacts and policy ideas, expanding our Symposia series.
The ideas outlined here are starting points for deeper exploration. The economic effects of AI are uncertain, and different scenarios will require tailored responses. Proactive engagement between researchers, policymakers, and the AI industry is essential to prepare for the future and ensure workers and communities benefit from AI's full potential.
These policy ideas have emerged from proposals and conversations with members of Anthropic's Economic Advisory Council, participants in our Economic Futures Symposia, and independent researchers. They do not necessarily represent Anthropic's official policy positions.
And this is the part most people miss... The future of our economy is intertwined with AI, and the choices we make today will shape tomorrow. Are you ready to join the conversation and help shape a future where AI benefits us all?
