I really liked Chemistry but ended up moving into software instead.
The dirty secret is that all factories are hard to build because nobody knows all the details to make them work.
It's that simple.
People bring operative knowledge to bear in the running of a factory. Over time, that knowledge becomes baked into the procedures, equipment, maintenance and people.
This is, in my opinion, something that everybody overlooks about nuclear. Power plants and factories need to evolve and optimize over time to be successful.
Nuclear plants get encased in amber and can't do that. I understand why people don't want to allow that. However, I really think that this inability to evolve will doom any large scale nuclear reactor design. Probably the only way that nuclear becomes successful is very small, semi-sealed power plants as the whole plant evolves at the manufacturing facility rather than at the site.
> I find it quite suggestive that most of the tech used to make the semiconductors is owned european companies.
This is hardly surprising. They're spinouts of the big conglomerates from the 1980s (ASML is from Philips, no?). These conglomerates didn't exist in Japan (maybe--MITI was funding the hell out of things in Japan in the early 1980s so my memory may be off), China, etc. back when this stuff was getting started and spun out.
TSMC is actually the anomaly. It took a very determined effort with a lot of money being shoveled around by the government combined with a disgruntled TI executive of Chinese background and all of his knowledge and contacts to put it all together.
I'm not sure this is the case. Chemistry and geology were both popular with hobbyists (albeit, it seemed to often be aristocratic hobbyists) during that period.
the competency is not allowed to spread. there's a thick shroud of secrecy around how all this sophisticated technology comes about.
This is also why semiconductors are so difficult.
Back in the early 20th century the nuclear stuff was secret so the nazis and then the russians would not get it.
Now semiconductors are also closely related to national security stuff (china and taiwan). I find it quite suggestive that most of the tech used to make the semiconductors is owned european companies.
Finally, I have a sensation that in the 18-19th century it was the chemical sciences that were similarly shrouded in secrecy of this sort.
There was a topic here on HN the other day about how there's so little popularization of chemistry... IMO, this is why, the legacy of secrecy so to guarantee competitive industrial advantages still casts its shadow.