lol i clicked on the article prepared to be angry that GK Chesterton wasn't mentioned and he was the first one mentioned. His collected works run to at least 37 volumes:
I happened to dip into Heretics recently and thought, if someone were to publish one of these essays as a blog post today, it would be decried as AI slop. This must be who the models learned to write from.
Most (but by no means all) prolific authors do, eventually. You see it especially in podcasts, many of them start to somewhat repeat after a few tens or hundreds of episodes.
Absolutely. When I was doing a more full-time analyst thing, I didn't reuse content wholesale but there was often very little point in rewriting an explanation of some basic concept with different words. (And, yes, as to volume, it really adds up if you're doing 1,000 words or so most weekdays.)
If you listen to hours and hours of someone who does "speeches without notes" you start to realize how they do it - they have "building blocks" that they piece together in different orders.
When I was doing more presentations than I do today, I didn't reuse presentations (unless it was part of some sort company roadshow). But, yes, I'd reuse bits and pieces--often tweaked--rather than each presentation being created from scratch.
About 15 volumes in I started to realize I’d read most of what he had to say; but it was still entertaining each time.
Given his size, perhaps he was the Largest Language Model.