he age of “generative” artificial intelligence has well and truly arrived. Openai’s chatbots, which use large-language-model (llm) technology, got the ball rolling in November. Now barely a day goes by without some mind-blowing advance. An ai-powered song featuring a fake “Drake” and “The Weeknd” recently shook the music industry. Programs which convert text to video are making fairly convincing content. Before long consumer products such as Expedia, Instacart and OpenTable will plug into Openai’s bots, allowing people to order food or book a holiday by typing text into a box. A recently leaked presentation, reportedly from a Google engineer, suggests the tech giant is worried about how easy it is for rivals to make progress. There is more to come—probably a lot more.

The development of ai raises profound questions. Perhaps most pressing, though, is a straightforward one. What does this mean for the economy? Many have grand expectations. New research by Goldman Sachs, a bank, suggests that “widespread ai adoption could eventually drive a 7% or almost $7trn increase in annual global gdp over a ten-year period.” Academic studies point to a three-percentage-point rise in annual labour-productivity growth in firms that adopt the technology, which would represent a huge uplift in incomes compounded over many years. A study published in 2021 by Tom Davidson of Open Philanthropy, a grantmaking outfit, puts a more than 10% chance on “explosive growth”—defined as increases in global output of more than 30% a year—sometime this century. A few economists, only half-jokingly, hold out the possibility of global incomes becoming infinite.