How fast have AI-producing sectors grown as a share of GDP?
Core argument: US AI-producing sectors' value added quadrupled from $278 billion to $1.13 trillion in constant 2017 dollars between 2010 and 2023, growing much faster than private nonfarm and manufacturing value added.
In the US, AI-producing sectors’ value added quadrupled from $278 billion (in constant 2017 dollars) to $1.13 trillion between 2010 and 2023, a rate much faster than those for private nonfarm and manufacturing value added. As a result, these sectors’ share in total US GDP increased from 2.4% in 2013 to 3.5% in 2023, with the data-processing sector nearly doubling its share in the same period. Meanwhile, the share of manufacturing declined by 1.5 percentage points. This fast growth of AI-producing sectors was driven by remarkable gains in labor productivity, with value added per employee in the data-processing sector growing about four times faster than that in the whole economy over the past 10 years his productivity growth was largely the result of elevated investment in physical capital and the complementarity of intermediate inputs, contrary to what was the case in computer systems design, in which labor and total factor productivity contributed significantly to output growth. Hence, the high output per employee in data centers, compared with that in other sectors, is the result of rapid capital accumulation, which has increased energy consumption as an intermediate input.



