Anthropic simulates the 2030 economy: AI enriches capital at the expense of labor

Anthropic has launched a simulator to measure the impact of AI on the U.S. economy by 2030. Its most radical scenario predicts a sharp acceleration in growth, but also a historic shift: capital capturing more income than labor. The article Anthropic simulates the 2030 economy: AI enriches capital at the expense of labor first appeared on Cryptoast.

The emergence of generative artificial intelligence promises to profoundly reshape global economic structures by the end of the decade. To anticipate these shifts, Anthropic has launched an interactive simulator called Econ Scenario Explorer. This tool allows for the modeling of the technology’s impact on the U.S. economy through 2030 by evaluating critical indicators such as GDP, unemployment rates, and income distribution. By analyzing specific job tasks according to the O*NET classification, this model highlights an unprecedented structural shift where automation redefines the landscape of the labor market.

The projections reveal contrasting trajectories, ranging from moderate growth to an extreme scenario marked by rapid acceleration. In the most ambitious hypothesis, U.S. GDP could surge by 32.4%, fueled by the massive automation of intellectual tasks. However, this macroeconomic prosperity would come with significant social instability, with unemployment estimated at 12%—a figure exceeding levels seen during major economic crises. This paradox raises a central concern: a wealthier economy does not automatically guarantee better income distribution for households.

The most striking finding lies in the transfer of value from labor to capital. While labor currently captures approximately 60% of generated income, Anthropic's extreme scenario suggests a complete reversal, where capital holders—companies, technological infrastructure, and AI models—would absorb over 54%. In practical terms, despite a rapidly growing economy, wages could stagnate or even decline by more than 10% for highly exposed knowledge-based professions. Additional wealth would then concentrate in the hands of asset owners, leaving knowledge workers in a state of increased precarity.

These perspectives shift societal concerns from the debate over employment to the ownership of digital means of production. The difficulty of retraining skilled workers for less automatable sectors underscores the labor market's inability to adjust quickly to this revolution. In the future, simply possessing technical expertise may no longer be enough to guarantee a stable standard of living. The fundamental question thus becomes one of asset access: in a world where AI generates a growing share of value-added, capital ownership appears to be the only hedge against the erosion of wage income.