"Like it or not, there's no guarantee of job security even for the plumbers and electricians and carpenters."

Well, to the extent that there is no guarantee of anything.

What make you think that improvements in technology in the next 30 years will enable machines to crawl under sinks or behind furniture; make estimates about corrosion, or wood grain, or wiring quality; manipulate small, complex items in tight spaces; explain the options (fix now but risk breakage later; replace now, but pay larger cost upfront) and answer questions, and do all this in a timely and friendly manner?

There is a large human factor to much manual labor that is hard to replicate with machines or remote operation.

Most everyone I know who ends up dealing with a non-local telephone support person hates it. Trying to get a decent explanation of why I should have my plumbing replaced rather than repaired from someone with poor localized language sounds skills sounds like a nightmare.

A better bet would be for, say, North Americans to operate remote repair devices for locations in North America (and likewise for other language-unified locales).

Perhaps a local, unskilled person operates a device to feed information back to a pool of technicians in some other location; you get the human element, but centralized technical smarts.

I believe something like that happens now with medical imaging. Technicians run the scanning machines, send the images off to who-knows-where for skilled readers to interpret, the results are moderated by a more-skilled supervisor, and sent back to the lab or hospital.

Still, I get the feeling that fixing bikes is part engineering, part art, and transmitting sufficient information over the wire for remote viewing may a hard problem. Like cooking, some things are just better handled by real live people.

Well, you seem to be dissagreeing, then agreeing with me, which is fine. I don't know where your 30 year deadline came from. Betting against the accuracy, efficiency, scalability and lower cost of our tools hasn't historically been a smart move, and I don't see why that would change. It's easy to envision a scenario where you could pay n times as much for an artisan plumber (or ... Bicycle Repairman!). But I bet most would opt for the cheap, standardized service of a tele-op'd Mario9000.

> What make you think that improvements in technology in the next 30 years will enable machines to crawl under sinks or behind furniture; make estimates about corrosion, or wood grain, or wiring quality; manipulate small, complex items in tight spaces;

A trip to your local Home Depot/Lowes shows that a huge amount of that home construction and repair that used to be a service has already become a user-installable product.

Many of the tight spaces for this stuff is a design decision that can change.

That doesn't really sound like an argument against skilled labor to me. Somebody needs to maintain those machines. Someone needs to set them up, plug them in, calibrate them, feed them raw materials, and generally keep them working.

As technology changes, the skills you need in order to rightly be called a "skilled" laborer change. In general, they seem to tend to require more education over time (although this might just be social; I'm not convinced that it really requires more study to become a programmer today than it did to become a good paleolithic hunter, it just seems easier to us because their training wasn't formalized and thus seems invisible). You can't expect one skillset to just carry you safely through the rest of your life without staying on top of changes.

Machines eliminate some jobs, but they inevitably create new ones: designing, maintaining, and operating the machines. In many cases they eliminate more jobs than they create, which is something we need to plan for, but I don't see any real chance that the machines are just going to make human beings obsolete -- in any avenue of work -- tomorrow.

As far as maintenance of telerobots, in the extreme case it's performed by other telerobots, manned by call centers full of "certified" gamers 10 or 10,000 miles away. Maybe not even call centers. Maybe employees are just perpetually on call, forced to "goggle in" periodically throughout the day to snake a drain in Brisbane or Bangladesh or Boise. In fact, they were probably goggled in already, playing U(niverse)oS(tarcraft) XXII, so they just alt-tab over.

Sounds crazy, I know, but it does seem to follow the general historical trend of alleviating manual labor from those of lower and lower socioeconomic status, while neatly sidestepping the problem of Strong AI with an ass-ton of artificial artificial intelligence.

The romantic, manly labor the article talks about requires the direct manipulation of physical reality with one's hands. The above scenario is skilled, certainly, but by no means manual.