Productivity
Rise of the machine
In the mid-1990s, technological progress slowed – bringing productivity in developed economies down with it. Today a generational shift may lead it back up.
Many economic forecasts display ‘productivity pessimism’, projecting relatively low or declining productivity growth into the indefinite future.
We might be at the inflection point of the J curve for AI.
Long-run economic growth is driven by productivity growth and the growth of the working population. So will AI lead a surge in productivity? In the past technological revolutions like electricity and the internet have undoubtedly had an important influence. However, the impact of these innovations occurred over a long period of time, say the BNY Mellon Investment Management (BNY Mellon IM) authors of the paper Tidal Forces: Dissecting the interest rate equation.
Shamik Dhar, chief economist at BNY Mellon IM, and Sebastian Vismara, senior economist, also point out that the yearly increase in productivity growth spurred by those innovations was relatively low.
Arguably AI could have a stronger yearly impact than those revolutions as it happens more quickly, the paper states. The paper’s authors believe this is the most likely scenario.
However, Dhar caveats this by noting ultimately this is an unknown that can’t accurately be predicted.
He says: “Assuming very little impact of generative AI on the economy and the world of work would require making just as heroic and hard to justify arguments. To the sceptics that point to weak productivity growth for more than a decade and argue that that is therefore likely to be the case going forward, we point out past productivity growth is a very poor predictor of future productivity growth.”
While there remain uncertainties, such as the computing power it requires for widescale adoption, Dhar is upbeat about AI’s influence on productivity. “It took about five years during the internet boom for productivity to rise following the increase in investment. Evidence suggests the time lags between the rise in AI-related investments and a subsequent boost to productivity could be even shorter. While barriers to adoption undoubtedly exist, they appear to be smaller than those faced during the electricity and internet booms.”
Many innovations follow a J-curve growth path. Initially, adoption starts slowly, but accelerates exponentially over time. In Tidal Forces: Dissecting the interest rate equation, the authors note the meteoric rise of ChatGPT, which has broken all previous records for the speed of technological adoption. This, says Dhar, highlights we might be at the inflection point of this curve for AI. “Based on parallels from historical trends, it's conceivable generative AI technologies could be widespread enough to elevate global productivity growth in just a handful of years.”