The differences in comparison to previous industrialization, like automated looms, trains, wood/metal lathes, automated phone systems, is that those machines all produced deterministic results. They could be understood by most mechanical or electrical engineers. You could see the obvious benefit of Ford's assembly line, and use that new fangled film technology to record what it looked like inside the new motor plants.
The LLMs have a lot of emergent properties. They are somewhat deterministic, as far as selecting the next set of possible tokens, but there is some randomness introduced so you're occasionally returned the token with 92% confidence over the 95% confidence, just to introduce some variability. But a new prompt that's as simple as "Are you sure?" can modify the context window so heavily you get entirely different results.
The new LLM era is very different, because these aren't deterministic machines that give you discrete results. You only have to do some image or video generation to know it takes several iterations to get what you want. Sure one person can animate their own 5~10 minutes shows now with just a couple of scanned in drawings, but they might have to render some things several times to get them right, and so they never know if the next episode will cost $200 or $600 in compute.