What you're not hearing about how AI is already powering a productivity supercycle

Productivity effects from AI are already starting to show definitively in the data. Since ChatGPT was introduced three years ago, growth in computer programming jobs -- previously a rapidly growing sector -- reversed. Yet programming output has continued to rise, thanks to soaring productivity growth as AI takes over mid-level programming tasks. This does not explain the post-pandemic productivity-led boom, though, because programmer jobs are only 1.5% of total jobs, and productivity must affect large-employment sectors to move the needle in the overall economy. The increased programming output is no doubt helping productivity in other sectors already. But a new productivity supercycle will be launched when AI brings productivity to the two largest employment sectors -- health care and education, fields that have been historically immune to productivity growth. These sectors rely on intelligence and verbal communication, the strong-points of AI. And they are low-hanging fruit for productivity gains, still at the beginning of their "S-curves." This will be the fuel for a coming productivity supercycle.