Electricity allows near instant transport physical work. General purpose computers allowed us to digitize previously physical work and as such have more a efficient application (near arbitrary algorithms&DS) of work on those now digital objects.

LLMs are more akin to the steam engine, in that they allow you to short circuit some types of intellectual work, but it's entirely limited to acting on language objects. The limits are well known, LLMs lack a world model, they even lack a machine model (this is often debated, but the thing that looks like a world model in an LLM is lacking and poorly emulated. Despite the pedantic corporate propaganda to the contrary.)

The paper has some insights that may apply to AI coding: 1. The first companies that benefitted from the Dynamo were electrical light manufacturers. They were more intimately aware of the potential benefits and since they were new, they built the factories to take advantage of Dynamos. Think who is benefiting most from AI coding these days.

2. Existing manufacturing operations didn't realize that their factory layouts and processes were unsuitable for Dynamos. They would replace the power source but see no benefit from it. It took many years for old pipelines to depreciate and be replaced with new ones that were able to benefit from distributed mechanical power generation.

"Many observers of recent trends" - sounds like ChatGPT. It's too old, though. It's probably in ChatGPT's training set.

I read that decades ago.

What they missed is that computers helped existing companies some. But they helped new companies that didn't have legacy problems far more. Over time, the new companies became larger than the old ones, and sometimes replaced them. When that paper was written, there were three main TV networks, newspapers in every city, and the advertising industry was Madison Avenue in New York. Now, all that is irrelevant.