Software outsourcing has been a thing for decades now and if anything seems past its peak as people find their programmers in Bangalore don't actually do as well as they hoped.
Which is because they went to the rock bottom bidder and that came at a price: you got the newbies, and the outsourcing company was training them on the job (ie _your_ project).
I heard stories that some companies rotated out their folks into higher paying projects (which existed) once they gained sufficient experience, replacing them with the next batch of newcomers - making the cheapest customers pay for "intro classes" (also known as amateur errors) during the entire project's lifetime. If true that would be a shining example of "you get what you pay for".
Hardware production outsourcing works well. You design the product, prototype it, create a specification that describes every necessary detail, and find a company to make your device. This engineering design process is expensive but does work well. Your onshore workers are the engineers, designers, and the like. Your offshore workers are the production staff.
Software production outsourcing works poorly. Creating a thorough enough specification for production is exactly as difficult as creating the program. You have all the same types of onshore workers, but there's no production staff!
I think the troubles caused by the decline of the old industries are largely due to the culture of investing your life in a company and only getting a salary out of it.
The software industry is closer to the creative industry where you try to extract as much as value possible and move on. Blockbusters and failures, agism, giants becoming outdated and loosing value like ageing milk and more culturally appropriate ones taking the scenes is not something that you can outsource.
Check out the Animation Department:
http://m.imdb.com/title/tt3120408/fullcredits/animation_depa...
The idea is that the industry will stay where the innovation happens and for innovation to happen well understood part of the work will be outsourced to a place that is willing to do the hard work for cheaper.
Maybe. But people have already tried this with off-shoring and it hasn't worked as well as people had hoped. I don't think the issue has been internet links being too slow.
I've outsourced some software development on Upwork and it's not as easy as it sounds to outsource work, you still need to be very engaged to make sure the people you hire are doing what you actually want. With a bearing factory, the process may be complicated, but evaluating quality is much simpler and completely standardized.
Software is not immune to labor market forces, but there is no clear outsourcing trend that I can see, so something would need to change for that trend to start.
how much and how well does the customer (in the US) know what they want in advance (as opposed to knowing what they want in reaction to the initial versions of the product)?
in other words: are US customers typically able to spec out the product well ahead of time -- or are they relying on a tight, iteration/feedback loop to help define the product?
Ed Yourdon said the same thing about thirty years ago (“Death of the American Programmer”). I’m still staying employed at a healthy salary. Turns out, apparently, when all things are considered, maybe the U. S. is the “cheapest place on Earth”.
Your #2 is predicated on Strong AI. At the current time there's no real evidence that that's forthcoming any time soon. (Hype notwithstanding.)
Currently, the most difficult problem with programming is understanding the problem. Absent Strong AI there's no shortcuts here.
... and while we're at it: IF the problem were to be solved -- even if by magic -- then I'd be the happiest person alive. The flaw in your thinking is that everything else would stay the same. It wouldn't. A culture with access to "solve-my-problem-X" on demand would be vastly different and it's not even clear if e.g. "money" or "job" would make any sense in such a world. (Ditto for any number of other concepts we take for granted, but which are ultimately driven by scarcity of resources or the fight against entropy.)
isn't that idea about 20 years old?
Also, shrewd programmers have been keeping the ball moving with a never ending stream of new ways (libraries/frameworks/platforms) to reinvent the wheel which has created a massive number of jobs for a long time. Some day this may come to an end as well, and I think the cloud might do it.
The universities are central to our economy. It's what defines this country.
And this will matter in any field, not just computer-science. It could be movies, agriculture, engineering, music, physics, photography, literature, fashion, medicine, business, whatever.. the universities help dominate an economy.
At the very least, we have to fight the right-wing conservative forces that are trying to cut our universities.
2) AI has nothing to do with this. Past outsourcing attempts have failed due to skill gaps and cultural differences.
3) The Internet is very good at superseding local culture with global Internet culture. Over time, past cultural differences will dissolve, especially considering that there is money involved.
4) The Internet is very good at making reference and learning materials available globally on the cheap as well as community support which can and likely will supersede a lot of what today's universities do. Over time I suspect that proximity to universities will have waning importance for all but the most specialized and advanced fields.
This is why American universities dominate.
2) Computer people complaining about efforts to repatriate factory work is pretty hilarious. You do realize where this whole industry is headed don't you? I mean no other industry is better suited for it, as tooling gets better, and links around the world get faster, it will be _the_ fastest sector to fly to the cheapest place on earth to do it.