A working paper from the Anthropic Institute models three economic scenarios for the U.S. economy through 2030, with an extreme case projecting 32.4% GDP growth alongside a 21.5% reduction in cognitive employment. The model assumes that 90% of AI-performed work replaces human effort rather than assisting it, leading to a shift in income distribution where capital share rises to 54.8% while knowledge-worker wages fall 11.5%. This represents a structural stress test on the assumption that AI-driven productivity gains automatically translate into broad-based labor market stability.
198 other tracked developments touching the same claims.