AI Could Add $10T to US GDP

Anthropic’s analysis is explicitly a scenario map rather than a forecast. It models the economy as millions of tasks based on the U.S. Department of Labor’s O*NET occupational taxonomy, allowing AI to leave tasks unchanged, augment them, automate them or create new tasks.
Anthropic said survey responses from more than 10,000 Americans clustered near its substantial scenario, rather than the extreme case, suggesting that the most disruptive outcome is not the central expectation reflected in public views.
In Anthropic’s substantial scenario, knowledge-worker wages could remain broadly flat while workers in other occupations experience stronger wage growth, indicating that AI’s labor-market effects may differ significantly across occupational groups.
Elon Musk has predicted that roughly one billion humanoid robots could be built within 10 years and that each robot might produce about five times as much as a human worker.
Musk also said AI could begin significantly affecting digital occupations by 2027, shifting human workers toward higher-level decision-making; the article additionally reported that SpaceX is investing in AI infrastructure, including data centers and electricity supplies, despite energy constraints.
Anthropic's economic modeling shows AI could add between $160 billion and $10 trillion to U.S. GDP by 2030, depending on how fast the technology advances and companies adopt it. The research maps three distinct futures: a modest scenario mirroring the internet's impact, a substantial middle case, and an extreme scenario where AI handles most knowledge work autonomously. Anthropic warns that even enormous economic gains may not help workers broadly, as wages for knowledge workers could stagnate while profits flow to capital owners.
Elon Musk went further, predicting AI combined with humanoid robots could more than double global GDP within a decade. Musk estimates one billion robots could be built in 10 years, each producing roughly five times what a human worker generates. This vision extends beyond software: SpaceX is already investing in AI data centers and electricity infrastructure to support the shift.
Anthropic built its forecast by modeling the U.S. economy as millions of individual tasks. Rather than guess, the team used the Department of Labor's O*NET database to see how AI might change each task—leaving it alone, improving it, automating it entirely, or creating new work. This task-by-task approach produced three scenarios with vastly different outcomes. The gradual case adds just 1.6% to GDP. The substantial case adds 8.3%. The extreme case—where AI runs most knowledge work—adds 32.4%.
Surveys of over 10,000 Americans clustered near the substantial middle scenario, not the extreme one. Anthropic says this suggests the most disruptive AI outcome is not what most people expect. The substantial scenario alone could add roughly $4 trillion to U.S. GDP by 2030, a massive shift but less apocalyptic than the highest estimate.
The wealth gains mask a harsh labor-market split. Anthropic projects that knowledge workers—lawyers, accountants, programmers, analysts—could see wages go flat or decline as AI copies their skills. Meanwhile, workers in construction, healthcare, and manual trades may see stronger wage growth because their jobs are harder to automate. This divergence means AI's benefits won't spread equally across the workforce.
Critically, Anthropic notes that higher GDP output may not lift worker prosperity at all. Income could shift dramatically from labor to capital. Owners of AI systems and robots would capture the gains while employees see little wage growth. This mirrors earlier tech booms where productivity soared but wages for many workers stalled.
Musk has called Anthropic's most extreme scenario—15% annual U.S. GDP growth—a certainty once humanoid robots hit mass production. He projects roughly one billion robots could be manufactured in a decade. Each robot would produce about five times the economic output of an average human worker, amplifying AI's impact beyond software alone. This pairing of AI brains with robotic bodies is Musk's formula for economic transformation.
Musk also predicts AI will meaningfully reshape digital work by 2027, pushing human workers toward strategy and high-level decisions rather than routine tasks. SpaceX is already preparing, investing heavily in AI infrastructure including data centers and power supplies. Musk is betting that companies which own both the AI and energy to run it will dominate the 2030s economy.
Publishers
12
Articles
2
Reach
14