Right, it seems the U.S. economy is trying to say, "making elementary machine tools is beneath us." The important bit is at the end: the U.S. is still a top buyer of machine tools (made elsewhere).
Is the same thing to happen with software?
When it becomes worth the time of the U.S. economy to produce basic machine tools again, they'll get to create new machine tools factories using all the latest technology: so it is probably good thing the "old way" is not still around hanging on by a thread.
Offshoring in software development has been around for a very long time. Most large US companies have a mix of onshore and offshore devs. The more mundane the software, and the tighter the financial macro-environment, the more the ratio shifts toward offshoring. This is the way the offshoring cycle has worked for a long time.
However, unlike hardware, software is about information and communication, and cultural context is very important. I have seen firsthand that non-US teams building software for US consumers often don't quite understand the reasoning behind the requirements and may lack polish around basic things like English. (The same is true, of course, in reverse, if US teams build for non-US audiences.) So I think it should be a little bit stickier.
You also cannot copy software design in the same way that you can copy, say, the design of a lathe. A lathe is a lathe, and as long as you've got the tools and materials, a lathe made in the US should not in theory be any different than one built in China. The same is not true of software.
> I have seen firsthand that non-US teams building software for US consumers often don't quite understand the reasoning behind the requirements and may lack polish around basic things like English
Something I'm curious about, have you seen this happen with countries culturally close to the US? Like teams based in Canada, Ireland, or the UK?
One of the biggest lies in cultural thinking is that countries like Ireland and the UK are "culturally close" to the USA. Sure they are close-er than say Italy or Japan, but anyone who thinks "close" means "similar enough that it doesn't matter" is in for a rude awakening.
In reality there are incredibly large gaps around even the most basic things like the meaning of words. As a very basic example: the meaning of the word "interesting" differs radically between cultures. Most US employees would think the boss is indeed interested when they describe an idea as "interesting" and may even bring it up again at some later date after more research. Meanwhile, someone from the UK means that it's the dumbest thing they've ever heard and they will be incredibly miffed if it is ever brought up again.
I've certainly experienced US-made software that was close to useless in the UK, demanding dates in some absurd middle-out format and expecting everyone to know what a "401k" was with no explanation, along with more minor bits of polish like saying "pound" but meaning a completely different symbol.
Hah, good point. If anything, software handling that kind of data might be better is smaller countries that need at international audience from the get-go!
Right. The US economy was well poised to tackle the massive task of computerizing humanity and, being a nascent technology stack with green field opportunities abounding, it was more profitable than continuing to produce machine tools. Meanwhile the Pax Americana has eliminated the risk premium to manufacturing overseas, so our entire economy has reconfigured itself around this task and opportunity of computerizing humanity. To our massive benefit, I think. IMO this is the main driver between diverging American incomes compared to the rest of the developed world.
As long as those manufacturing centers are part of US allies (Japan, Korea, Europe, Australia, Taiwan), or at least trustworthy neutrals (Vietnam, India), I'm happy.
But some manufacturing, in fact a large amount, is Chinese. And I'm not convinced they've got our best interests at heart. Either China needs to calm down over their "Wolf Warrior" style, or we need to cut back on providing benefits to an obviously and increasingly antagonistic power.
---------
The other part, with respect to Taiwan + China, is that we must defend Taiwan as they are a critical source of advanced-materials (ie: computer chips) to us. Yes, its more efficient to have Taiwan centralize production, but it does come at a cost. We need to be ready to defend Taiwan and keep it safe if we are to continue to build computers and phones out of Taiwan-only parts.
The entire economy is configured around making quarterly profits for shareholders. Any good that happens to come out for humanity is purely incidental. The US economy is not some fairytale hero.
I have a theory that one of the major challenges facing manufacturing in the US is actually the opposite.
A manufacturing company in the US is competing for smart, numerate STEM graduates with the likes of Google who offer graduates $180,000 with zero experience (or so I'm told)
But they're also competing with manufacturers in the far east, where $40,000 is a great salary, for someone with several years of experience.
I think that’s right, and I think there’s a third aspect at play, which is that culturally the US doesn’t value jobs in manufacturing, machining, engineering, etc. as highly as it once did (after WW2, say) or as highly as it values finance and tech jobs today. I’m not sure how easily that could change.
Will “all the latest technology” be available? The expertise the US once had in this area is largely gone, and won’t come back immediately. I think there’s real reasons to be concerned that the US will fall behind countries like China in areas like innovation because of this self-inflicted brain drain.
Right now there are mandates to build out the manufacturing base needed for military efforts.
They know the tech sent off shore will, for the most part, remain off shore.
Additive manufacturing, an example of "latest tech", is a primary build up target here. It is new and expertise is being created as the tech itself is.
HYBRID DED is going to be the answer to many of those scenarios.
Basically, we couple subtractive with additive and maximize the benefits of both. Done right the risks are greatly reduced as are costs compared to just doing it the legacy way.
Is the same thing to happen with software?
When it becomes worth the time of the U.S. economy to produce basic machine tools again, they'll get to create new machine tools factories using all the latest technology: so it is probably good thing the "old way" is not still around hanging on by a thread.
The market naturally is culling technical debt.