It's not that existing jobs will always be automated, we'll come up with tools that completely automate several layers of work.
It's a different view than all current jobs will be automated. Automation can be used to remove several verticals and improve efficiency by magnitudes in the process.
People are very good at bricklaying, and are paid a reasonable rate by the hour. No one has made a machine that can beat that feature set.
edit: more great features of human brickies that robots are terrible at: very small working volume; very small and agile footprint; safe for other humans to work near; half-day built-in power supply.
Where I live in Indiana an average single family home is like 200k. No one is paying 50k for brick on that, so vinyl siding you go.
It's brand new, give it a minute.
The problems with new building tech are twofold:
First of all, for most people a house is the biggest investment they'll ever make, so they're not really keen to gamble. Oh, they'll make minor gambles - but a complete new way of building houses is a tough sell.
Second of all, building a house involves so many different costs that even a big saving in one area is only a drop in the ocean on the total project cost. If I'm building a brick-and-block house, the total cost of materials and labour might be $180,000 with brickwork comprising 10% of that. That's nothing to be sniffed at - but no matter how great and cheap your brickwork replacement is, someone taking a risk and using your new tech can't possible save more than 10%.
I can walk down my street in Northern New England and see brick buildings that are over 100 years old in a harsh climate, and that were reasonably affordable when they were built. The shapes and contours are usable and attractive to most people. They are able to be remodeled and repaired using very basic technology.
There are reasons that Dymaxion Homes weren't popular.
The only problem is they seem to attract tornados.