The Simple Macroeconomics of AI

2024

Open publication workspace · Sign in to read the full PDF.

AI-generated summary

1) This paper provides a grounded macroeconomic evaluation of generative AI, demonstrating that its medium-term impact on productivity and GDP will likely be much more modest than widely claimed.
2) What's inside:
- Applies Hulten’s theorem and task-based modeling to estimate that AI will increase total factor productivity (TFP) by no more than 0.71% (or 0.55% when factoring in hard-to-learn tasks) over a 10-year horizon.
- Analyzes wage and inequality implications, showing that while AI exposure is evenly distributed across demographics, it may slightly increase between-group inequality and widen the capital-labor income gap.
- Incorporates welfare considerations for AI-generated "bad tasks" (such as misinformation and online manipulation), illustrating how revenue gains can mask negative societal impacts.
3) Tags: artificial intelligence, macroeconomics, productivity, inequality, automation

Check the original publication for accuracy and context.