"On top of this, common economic worries are less fearsome than they first sound. One concern is that artificial-intelligence investment spending, especially on data centres, is the only thing keeping the party going, spelling disaster _if investors cool on the tech_.^1 This argument is strongest in America, where investment in information-processing equipment and software (IPES) accounted for 40% or so of growth in real GDP over the past year. Yet, at a minimum, two-thirds of IPES has _nothing to do with AI_.^2 The data include, for instance, a business buying a computer. In addition, outside America there is _no evidence whatsoever that IT is driving growth_."
"A second concern relates to jobs. Employment growth in America has slowed, raising fears of _AI-induced unemployment_.^3 Not so fast. A new study by the Yale Budget Lab finds that "the broader labour market [beyond the so-called "tech" industry] has not experienced a discernible disruption since ChatGPT's release." Outside America there is little evidence of an employment slowdown. In the first half of the year other OECD countries added 3m jobs, in line with the norm before the covid-19 pandemic."
Perhaps the so-called "tech industry" and the endless autocomplete-powered gibberish hype they sponsor is not an accurate portrayal of the American or world economy
1. That is, growth may continue if Ai investement declines
2. Growth attributable to "IPES" is only 40% compared to 60% attributable to non-IPES, and 2/3 of IPES has nothing to do with "AI". Further, the definition of "IPES" is so broad as to include a business purchasing a computer
3. Every day, "AI" hype beats web users over the head with this meme that "Ai" is going to cause mass unemployment. There is no evidence of this causing an employment slowdown in the US post-ChatGPT and outside the US in the OECD there has been no employment slowdown post-ChatGPT. But who needs evidence, which usually requires work to gather, when daily speculation is free and requires zero effort
zahlman•3mo ago
The quoted portion appears to contradict you, and you offer no reason why it should be incorrect.
zahlman•3mo ago
> Meanwhile, gung-ho fiscal policy, especially in America, is stoking demand.
I assume they're referring to the spending in the "big beautiful bill"?
1vuio0pswjnm7•3mo ago
"A second concern relates to jobs. Employment growth in America has slowed, raising fears of _AI-induced unemployment_.^3 Not so fast. A new study by the Yale Budget Lab finds that "the broader labour market [beyond the so-called "tech" industry] has not experienced a discernible disruption since ChatGPT's release." Outside America there is little evidence of an employment slowdown. In the first half of the year other OECD countries added 3m jobs, in line with the norm before the covid-19 pandemic."
Perhaps the so-called "tech industry" and the endless autocomplete-powered gibberish hype they sponsor is not an accurate portrayal of the American or world economy
1. That is, growth may continue if Ai investement declines
2. Growth attributable to "IPES" is only 40% compared to 60% attributable to non-IPES, and 2/3 of IPES has nothing to do with "AI". Further, the definition of "IPES" is so broad as to include a business purchasing a computer
3. Every day, "AI" hype beats web users over the head with this meme that "Ai" is going to cause mass unemployment. There is no evidence of this causing an employment slowdown in the US post-ChatGPT and outside the US in the OECD there has been no employment slowdown post-ChatGPT. But who needs evidence, which usually requires work to gather, when daily speculation is free and requires zero effort
zahlman•3mo ago