There's a startup building prefabricated house using robots and cheap labor in a factory, with very little professional work on site.

But on the other hand, yesterday we had an article here about visual coding, and according to some it can build very complex application with much higher productivity(and maybe the work becomes accessible to business analysts and domain experts, maybe).

So it's hard to tell these days.

>There's a startup building prefabricated house using robots and cheap labor in a factory,

Not my area but various companies have been trying to do this sort of thing forever. I can remember reading stories about partially pre-manufactured modular homes decades ago and it's never panned out. It seems to be one of those ideas that makes a lot of sense on paper but people don't want it.

I grew up in a Sears & Roebuck home, as were a good portion of those on my block. The Wikipedia article on the subject says that there were 70,000 of them built[1]. So I see this as more of something that has come and gone, probably a few times, and we are back on the up-swing.

The Sears & Roebuck were partially assembled (think wall by wall), and then those parts were hoisted off the freight trains and nailed together. There was still a lot of work that needed to be done (often by professionals), but it really cut down on the (especially local) costs.

[1] https://en.wikipedia.org/wiki/Sears_Catalog_Home

Katerra?

They recently raised $865m lead by SoftBank's vision fund [0].

[0] https://www.nytimes.com/2018/01/24/business/dealbook/katerra...

I think “the last mile” of this process is going to be the hardest to automate.

I see early childhood and primary school teaching going later.

Quick! someone write a short story about AIs who hire a professional human.

You don't automate, you make things more efficient. The building industry is very good at bringing to market products that increase productivity by double or more.

So people will have to build more houses for the same money.

In many cases, you'll end up making something that works using simple relays way more complex, so you'll still need some kind of technician to troubleshoot the equipment when it fails. The "advances" in equipment technology is mostly for sales brochures but in many cases the increased complexity creates more failure points and simply requires a different type of technician to troubleshoot.

I'm talking about things like PEX, which drastically simplify and reduce the time needed to plumb a house. The construction industry as a whole seems to be making a huge investment in time-saving innovations.