Anthropic Tracks AI's Economic Impact Amid Job Loss Concerns

The Economic Implications of Generative AI: Anthropic's New Initiative

Silicon Valley has extensively discussed the potential of generative AI to stimulate the creation of novel career opportunities and drive economic growth, exemplified by the emerging concept of the solo unicorn startup. Financial institutions and analysts have highlighted AI’s capacity to enhance Gross Domestic Product. However, these benefits are not anticipated to be distributed evenly, particularly given widespread expectations of job displacement due to AI.

Anthropic Launches the Economic Futures Program

In response to these concerns, Anthropic unveiled its Economic Futures Program on Friday. This new initiative is designed to support research into the effects of AI on both the labor market and the global economy. It also aims to formulate policy recommendations to proactively address the impending changes.

“There is considerable inquiry regarding the economic consequences of AI, encompassing both its positive and negative aspects,” stated Sarah Heck, head of policy programs and partnerships at Anthropic, in an interview with TechCrunch. “It is crucial to base these discussions on empirical evidence, avoiding preconceived notions or biases about future developments.”

Potential for Significant Economic Disruption

Dario Amodei, CEO of Anthropic, has already offered his perspective on the potential economic impact of AI. He predicted in May that AI could eliminate up to half of all entry-level white-collar positions, potentially leading to an unemployment rate as high as 20% within the next one to five years.

When questioned about whether a primary objective of the Economic Futures Program was to investigate methods for mitigating AI-induced job losses, Heck responded with caution. She acknowledged that the transformative changes brought about by AI could be “both advantageous and detrimental.”

“Our main focus is to accurately understand what is occurring,” she explained. “Should job losses materialize, we should bring together a diverse group of experts to explore mitigation strategies. Conversely, if substantial GDP expansion occurs, that’s positive as well. We should then engage policymakers to determine the appropriate course of action. This will not be a uniform outcome.”

Building on the Economic Index

This program expands upon Anthropic’s existing Economic Index, which was launched in February. The index provides open access to aggregated, anonymized data, enabling analysis of AI’s effects on labor markets and the economy over time – a practice that distinguishes it from many competitors who keep such data proprietary.

Three Key Areas of Focus

The program will concentrate on three principal areas:

• Providing grants to researchers studying AI’s impact on labor, productivity, and value creation.
• Establishing forums for the development and evaluation of policy proposals to prepare for AI’s economic consequences.
• Constructing datasets to monitor AI’s economic utilization and impact.

Initial Program Actions

Anthropic is initiating the program with several concrete steps.

The company is currently accepting applications for rapid grants, offering up to $50,000 for “empirical research on AI’s economic impacts,” as well as for evidence-based policy proposals. These proposals will be presented at Anthropic-hosted symposia events in Washington, D.C., and Europe this fall.

Anthropic is also actively seeking partnerships with independent research institutions and will provide collaborators with Claude API credits and other resources to facilitate their research.

Grant and Symposium Requirements

Regarding the grants, Heck emphasized that Anthropic is seeking individuals, academics, or teams capable of producing high-quality data within a relatively short timeframe.

“We aim to complete the research within six months,” she stated. “Peer review is not necessarily a requirement.”

For the symposia, Anthropic desires policy ideas from a broad spectrum of backgrounds and intellectual viewpoints, according to Heck. She clarified that policy proposals should extend “beyond labor considerations.”

“We want to gain a deeper understanding of the transitions taking place,” she said. “How are workflows evolving? What new jobs are being created that were previously unimaginable? How are certain skills retaining their value while others are becoming obsolete?”

Exploring Broader Economic Effects

Heck also indicated that Anthropic intends to investigate the effects of AI on fiscal policy. For instance, they will examine the implications of a significant shift in how businesses perceive value creation.

“We want to broaden the scope of inquiry to encompass a wider range of topics,” Heck explained. “Labor is undoubtedly one aspect, but it represents a much larger landscape.”

Comparison with OpenAI’s Approach

OpenAI, a competitor to Anthropic, released its own Economic Blueprint in January. This blueprint focuses more on facilitating public adoption of AI tools, developing robust AI infrastructure, and establishing “AI economic zones” with streamlined regulations to encourage investment. While OpenAI’s Stargate project, a collaboration with Oracle and SoftBank to build data centers across the U.S., is expected to generate thousands of construction jobs, OpenAI’s blueprint does not directly address AI-related job losses.

However, OpenAI’s blueprint does propose frameworks for government involvement in supply chain training, investment in AI literacy, support for regional training programs, and expanded access to computing resources for public universities to cultivate local AI-skilled workforces.

A Growing Trend in the Tech Industry

Anthropic’s economic impact program reflects a growing trend among some technology companies to position themselves as part of the solution to the disruption they are contributing to—driven by concerns about reputation, genuine altruism, or a combination of both. As an example, Lyft launched a forum on Thursday to gather feedback from human drivers as it begins integrating robotaxis into its service.