> Interesting to compare the tech companies of last century with those of this era.
The 'tech companies' (equivalent) of the past tended to deal with physical goods, and so needed physical means of scaling to become as large as they did.
More recent tech companies are often software goods and services, where physical means of scaling may not be as important (though see perhaps with Moore and Dennard). Though those which deal with physical stuff do seem to have higher counts; see below.
> The number of people benefiting from an enterprise has shrunk considerably, with the benefits accruing more tightly within an already wealthy class.
Not sure if this is completely accurate. Kodak topped out at 145,000 employees in 1988:
Many tech companies still deal with physical goods and employ enough people to fill a city - but no longer in the US and Europe, generally. Whether we want them back is up to us.
Ever been to Folsom, Chandler, Olathe, Huntsville, Quad Cities, Peoria, Irvine, San Diego, North Chicago, East Liberty, Taylor, etc?
We have seen a decrease in companies that work on physical engineering goods, but plenty still remain.
The difference is, engineering physical goods has been increasingly automated and skilled for decades now (robotic manufacturing, additive manufacturing, CAD tooling, etc has been a thing for decades now), so you will need at least an associates but often a bachelors degrees in a STEM field to climb the ladder in Automtotive, Aerospace, Electronics, Pharma, Chemicals, and Defense manufacturing.
Even the "MBAs" they hire in LDPs are required to have technical undergrad degrees before joining an MBA.
I don't like the fact that stores like Fry's don't exist anymore, but electronics is a much more niche hobby now than it was 30/40/50 years ago and is much less hobbyist friendly - building something with an 8088 might be fun, but not as much anymore when you can purchase a STM32 for a fraction of what they 8088 was 30 years ago. Coding has becoming a very prevalent hobby now, and access to hobbyist boards like Arduinos and Raspberry Pis has become much more democratized.
I lived in one of these cities. I don't even understand what you're arguing against. The US has highly automated manufacturing because that's what makes sense with the current incentives. Other parts of the world still have factory cities.
The US can align incentives to encourage different types of manufacturing of if it wants. I didn't say anything about whether it's a good idea.
US has highly automated manufacturing but only in a few industries where they're still competitive or have trade protections.
Though I feel like generally the narrative being spun for folks is that factory jobs largely got automated, which is true only in the fact that the remaining factory jobs got automated. The vast majority of them went overseas because it was cheaper. Otherwise the country wouldn't be filled with rusting factory towns and you'd still be able to find lots of household goods made in the US.
> Other parts of the world still have factory cities
I mean, factory cities in other countries are also increasingly being automated as well.
Look at what happened to Wolfsburg in Germany as well as the investments in robotic and additive manufacturing in China leading to around 20 million factory jobs being shed from 2013 to 2023 [0]. If an Indian auto manufacturers like Maruti Suzuki themselves had a 1 robot for 4 employee ratio by 2017 [1] in a country where labor (skilled and unskilled) is cheap, it shows where the tide is turning globally for factory cities.
Globally, manufacturing (especially in electronics, chemicals, pharma, automotive, and aerospace) has become much more automated because the cost barriers for automation have fallen significantly. For example, a robotic arm like that which Kuka or FANUC manufactures costs around $30-40k to install, whereas 30 years ago an industrial robot would have costed around $65-115k in 2003 dollars [2]
The era of single factory cities is coming to a close globally. Industries are overwhelmingly clustering in hubs in order to take full advantage of supply chains and vendor ecosystems.
I'm fine with all of this. Factory hubs are the new factory cities. You still get to employ thousands who don't necessarily have a PhD in computer science.
> I don't like the fact that stores like Fry's don't exist anymore, but electronics is a much more niche hobby now than it was 30/40/50 years ago and is much less hobbyist friendly - building something with an 8088 might be fun, but not as much anymore when you can purchase a STM32 for a fraction of what they 8088 was 30 years ago. Coding has becoming a very prevalent hobby now, and access to hobbyist boards like Arduinos and Raspberry Pis has become much more democratized.
I don't think it's so much that electronics is more niche of a hobby but instead of going to Frys or Radio Shack people can click a button and order from Amazon or the Arduino store or anywhere else. There's no hoping a shop has a part you need. You just pick it and are done.
I still don't understand why San Diego-Tijuana isn't in principle as strong in design+manufacturing as the Hong Kong-Shenzhen axis. (What needs to happen in order to make that so?)
> Even the "MBAs" they hire in LDPs are required to have technical undergrad degrees before joining an MBA.
It frightens me the quotes you put around "MBAs" like that, those are the best type of MBAs. In general the cadre of people who go directly or an MBA with zero industrial exposure have very different aims than the ones that do (also their funding sources are very different).
> I still don't understand why San Diego-Tijuana isn't in principle as strong in design+manufacturing as the Hong Kong-Shenzhen axis
It is in Medical Devices and Defense Tech.
The San Diego and Tijuana EDCs are closely integrated and use a single-window system so that you as a company doing R&D in SD and prototyping and manufacturing in Tijuana are dealing with a single entity.
> It frightens me the quotes you put around "MBAs" like that
Because on HN, people use "MBA" to designate CorpDev, FP&A, CorpStrat, GM, and other management and revenue ownership roles.
> those are the best type of MBAs
I agree, and employers agree as well.
There's a reason alumni from Wharton SF, Haas PTMBA, Stanford HCP MS&E, MIT's Sloan+LGO Fellowship, CMU Tepper's PTMBA, UW Foster's EMBA, UT McCombs' Hildebrand MBA, and UCLA Anderson's FlexMBA are overrepresented in engineering industries.
Employers tend to sponsor high leadership potential employees to attend these programs in return for them climbing the ladder internally.
An entire generation of Cybersecurity and Enterprise SaaS CPOs and CEOs are products of this initiative at Cisco and PANW back in the 2000s
I was a Folsom resident for 15 years. If you’re referring to physical goods, Folsom didn’t produce any. Perhaps you thought Intel site there was a fab? (FYI: Folsom was a small post-goldrush dwindling mining town that turned into a suburban hellhole thanks to Intel: The same strip malls on every block and beige-box sprawl as far as the eye could see. It’s a great example of how rapid scaling culturally starves humans in the long run ... unless you like highschool theater and boomer cover bands.)
What I mean is Folsom is a city where research into a physical engineering field continues, and primarily for a single employer (Intel), though that remains to be seen with the current rightsizing going on.
> The same strip malls on every block and beige-box sprawl as far as the eye could see
Plenty of industrial cities are similar. And as I mentioned in the previous response, manufacturing is much more automated and skilled today - it's increasingly a white collar job in nature.
It's also changing political and social landscape though, Cadbury's (though now Kraft-owned) still makes physical chocolate but not from an employee-only model village. Ditto Clark's Village (shoes).
(They both happen to be Quaker I believe, but I mean the point more generally.)
Heh, I was musing on something I'd read recently that had Kodak as a case study. It was compared with Instagram which had about 10-15 staff when it was bought for USD 1bn. WhatsApp similarly small.
I hear tech can be big employers, maybe I'm overselling my point a bit there. That said, the trend is very much towards smaller operations, and a large headcount is not at all required for large money.
My overall point is that profits that at one time would require a town, or be a major part of a city's economy, can be made with a small office's worth of staff.
That's a challenging comparison, though, as Instagram never had to do fundamental research. They wrote an app. One that has to scale to enormous traffic, to be sure, but that came after they were a 10-15 person company. Meanwhile, Kodak was inventing lens technology and film chemistry and manufacturing processes and distribution channels.
Similarly, Ford has more employees than Gran Turismo, but they're not in the same industry.
To be fair maybe a better comparison would also be all the people maintaining the physical infrastructure Instagram was using (no idea what Cloud provider they originated on)
> My overall point is that profits that at one time would require a town, or be a major part of a city's economy, can be made with a small office's worth of staff.
This has been true for most industries even 'with-in themselves': it's called productivity growth.
The number of employees (or man-hours) needed to create (say) a tonne of steel has dropped a lot, so where previously you had 'steel towns', now a plant may just have a very few (and produce more tonnage than they've ever done).
But they're part of its stack. It wouldn't exist without it. And vice versa, AWS wouldn't have as many employees without companies like Instagram.
Kodak required so many because it owned nearly the whole stack.
But that said, we can't just add the number of AWS employees to the number of Instagram employees, because we have to prorate it by the total number of companies that use AWS (or whatever).
The 'tech companies' (equivalent) of the past tended to deal with physical goods, and so needed physical means of scaling to become as large as they did.
More recent tech companies are often software goods and services, where physical means of scaling may not be as important (though see perhaps with Moore and Dennard). Though those which deal with physical stuff do seem to have higher counts; see below.
> The number of people benefiting from an enterprise has shrunk considerably, with the benefits accruing more tightly within an already wealthy class.
Not sure if this is completely accurate. Kodak topped out at 145,000 employees in 1988:
* https://rbj.net/2017/09/13/kodaks-decades-of-decline/
Apple has 164,000; Microsoft has 228,000; Amazon has 1,610,000.
* https://en.wikipedia.org/wiki/List_of_largest_employers