What I came away with was first, the science of plant automation is far, far more advanced than computer science :-) Second, they've eliminated an incredible about of manual labor already.
Using factory automation as a model, it's interesting to think how software could do the same thing to the marketplace.
I agree with @nl, as long as the worker is there voluntarily, I don't see a problem with a market existing for sweating the small details. The upstream market competes by optimizing their BOM, so if the zipper isn't worth the cost, it will phase out.
To argue that it shouldn't be worth the cost because the job seems to boring for you is odd. There are many people who would be very happy to have such a job. If and when there is a future where there does not exist a large number of people would be overqualified for such a job, let me know.
That oversimplifies what I'm saying.
A simplified version is this: It would be better if everyone was able to strive for self-actualisation[1] without the economic constraints that require people to work. In the absence of that, and while there are plenty of people at low levels on Maslow's needs hierarchy[2] who need to work, then there is nothing wrong in wishing that the jobs they did provided some level of value beyond being a replaceable cog in a machine.
[1] http://en.wikipedia.org/wiki/Self-actualization
[2] http://en.wikipedia.org/wiki/Maslow%27s_hierarchy_of_needs
It's not that simple.
While I don't preclude that is a possible outcome, if you read the entire article you see that the fact that smooth zippers are more expensive means they can only be used on more expensive items, even though they don't give any practical increase in quality in terms of reliability.
If those smooth zippers were as cheap to produce as the ones with bumps, then it's possible (likely even? - after all, producers always want to differentiate their good in some way) the decreased cost of production would be invested in adding features that actually do something. One defensible possibility is "mass customization" type production, which relies on each piece being deliberately different in some way. I raise that possibility because most would consider that type of work to be "better" than manually threading zippers, and there could be more people employed.
Clearly there are a lot of things that could happen if both forms of zippers were as cheap to produce. My point is that it is difficult to say for sure the worker would be on the street with no job.
[1] It's more likely that the worker is actually female: Chinese factories generally prefer to hire women. It is only comparatively recently that the ratio pf production line workers has approached 50/50, and as recently as 2006 there were 14 men for every 100 women working on production lines in China. See http://www.ft.com/intl/cms/s/0/c8948f1e-2653-11e3-8ef6-00144...