AI, productivity, & geopolitics
Long-time readers of these notes may recall my argument in 2020 that a productivity renaissance is quite likely over the next decade, after a sustained period of declining productivity growth since the 1970s across advanced economies.
The basic argument was that the pandemic shock led to substantially increased investments in capital and technology, changed business models (digital, virtual), as well as introduced more flexible working models. And tight labour markets across advanced economies create incentives for labour productivity enhancing investments.
However, the productivity data available so far are not that encouraging. Data through to Q1 show a flat trend over the past several quarters after an initial productivity recovery/surge in the early stages of Covid. And the near-term projections from the OECD, IMF, Conference Board, and others also suggest a muted productivity growth outlook.
Believe (some of) the hype
Despite this, I remain upbeat on the productivity outlook for three reasons. First, the strengthening of inputs to the productivity process (capital investment, new business models) is real; and so I am prepared give the hypothesis some more time. And the labour productivity data over the past several quarters is unusually noisy because of the rapid post-Covid recovery in hours worked, with tightening labour markets and a rebound in labour intensive services sectors.
Second, recent geopolitical dynamics and the emergence of a ‘wartime economy’ is supporting substantially increased investments in industrial and innovation policy. This will have positive effects on productivity over time, particularly given that many of these investments are being made into advanced technologies, green energy, infrastructure, and so on.
Although growing global economic fragmentation will drag on productivity, these costs will be offset by increased investment. And the rapid demographic reversal sharpens the incentives for productivity-enhancing investment. (I didn’t find the Economist’s recent assessment that aging populations will weaken innovation capability fully persuasive.)
The third factor that makes me positive on the productivity outlook is the rapid development of generative AI. This has burst into public awareness over the past 6 months: ChatGPT reached 100m monthly active users in 2 months; and share prices for AI-exposed firms have exploded (Nvidia is up by >3x in the past 8 months, reaching a market cap of >US$1 trillion). And a 4-week old Paris-based AI start-up raised €105m earlier this month…
The full note is available at: https://davidskilling.substack.com/p/ai-productivity-and-geopolitics