logoalt Hacker News

crossbodytoday at 5:46 PM1 replyview on HN

LLM hallucinations are still a thing for ultra niche topics. Not a problem for topics that have sizeable wiki pages, like Baumol Effect. Here is the first paragraph from wiki: "...tendency for wages in jobs that have experienced little or no increase in labor productivity to rise in response to rising wages in other jobs that did experience high productivity growth"


Replies

PaulDavisThe1sttoday at 6:35 PM

The problem with this summary is that it's not actually what Baumol & Bowen were really focused on in their original paper.

It is what most people nowadays connect with "Baumol's cost disease", but in their paper, the way in which rising productivity sectors cause wage increases in stable productivity sectors was more of a detail than the central part of their thesis. The core part was the observation that certain kinds of economic activity cannot reduce costs through productivity gains, while others can; the wage connection between them was, well, not an afterthought, but more of a consequence of the very specific sort of economic system we live in. One could imagine a society with different ways of distributing resources to labor that didn't really have this feature, and yet the same sectors of this imaginary society's would still suffer from "Baumol's cost disease".