Symbol s

The passage around this formula

When applied to the Harvard BCG findings, Acemoglu’s macroeconomic framework reveals why the study’s 40% performance gains cannot be translated linearly into organizational throughput: $\pi_s = \frac{c_L(s) - [c_{\text{draft}}(s) + c_{\text{audit}}(s)]}{c_L(s)} \ll 0.40$. Using rigorous baseline accounting across the United States economy, Acemoglu estimates that generative artificial intelligence will affect no more than 4.6% of total tasks over a ten-year horizon, producing an aggregate TFP increase of less than 0.71% cumulatively over a decade (approximately 0.07% annually) [ 2 ] . The breathless projections of 40% corporate workforce expansion derived from micro-task consulting studies collapse when subjected to rigorous general-equilibrium accounting.

Read this part in the article →