Uncertainty about how artificial intelligence will reshape the job market has dominated boardroom discussions, career planning, and economic policy debates worldwide. Yet, concrete data on the scale of these changes remains scarce. Anthropic, the company behind the Claude system, has attempted to address this gap with a study presenting three distinct projections for the US economy by 2030, ranging from moderate impact to radical transformation.
This analysis is not impartial. As a major beneficiary of AI’s expansion, Anthropic’s report serves both as a serious forecast and as an advocate for its commercial interests. Nonetheless, the document offers a rare quantification—in dollars and percentages—of how varying levels of AI adoption could affect wages, growth, and employment.
Modest scenario: an impact comparable to the internet
In the most conservative scenario, AI would boost US GDP by 1.6%, reaching $34.1 trillion—an effect similar to that of the internet over several decades. In this case, most professions would continue to exist much as they do today, with AI primarily serving as a support tool rather than replacing human labor.
Substantial scenario: half of intellectual work automated
In the intermediate scenario, AI would autonomously perform half of all intellectual tasks, driving GDP growth by 8.3% to $36.3 trillion. Economic growth would double its usual pace, but wages for intellectual professions would remain relatively stable, with no sharp income declines for those who stay employed.
Extreme scenario: accelerated growth and record unemployment
In the most radical scenario, the US economy would grow by 15% annually, doubling in size every 4.5 years. This progress would be fueled by a cycle of recursive self-improvement, where AI itself would help develop more advanced versions of itself. In this case, unemployment would surpass levels historically linked to recessions, and the share of wages in GDP would drop from 60% to 45.2%.
Public perception of the scenarios
An Anthropic survey of over 10,000 Americans found that most respondents expect developments to align closely with the substantial scenario. However, about 10% of participants believe the labor market transformation will be rapid and radical, matching the extreme scenario.
In summary, Anthropic’s projections offer a range of possibilities to help governments, businesses, and workers prepare for the impacts of artificial intelligence in the coming decades.
