In a widely debated NBER working paper, MIT economist Daron Acemoglu (2024 Nobel laureate in Economics) challenged the prevailing narrative of massive AI-driven economic transformation:
- AI will likely increase US GDP by only 0.5-1.5% over the next decade — far below predictions of 15-25% by Goldman Sachs and McKinsey
- Only about 4.6% of worker tasks will be meaningfully automated by AI in the next 10 years
- The gap between "technically automatable" and "actually automated" is large — due to cost, reliability, and organizational constraints
- AI benefits may accrue primarily to capital owners, potentially increasing inequality rather than broadly shared prosperity
The Sobering Counter-narrative
Acemoglu's analysis doesn't deny AI's potential — he argues the timeline and distribution of benefits are being drastically overstated. This matters because over-hyped expectations can lead to misallocated investment and inadequate safety precautions.
Source
Acemoglu, D. (2024). The Simple Macroeconomics of AI. NBER Working Paper No. 32487.