Human labour is extremely and rapidly adaptable. Specialised labour still needs to repair the robots, and there may be a handful of 'easy' tasks where the cost of automating is higher than the cost of employment. This is already the case if you see a mass produced manufacturing line.

The real impending issue is that automation is only possible with access to capital, and whether it lowers average cost.

Those robots are often very specialised and very expensive. The problem is that they are causing capital to have increasing power and productivity, leading towards it being more efficient and profitable to have a smaller number of companies who are able to make that investment, and other just die out (a large barrier to competitiveness). This happened in the 1900s with cars - GM/Ford ect bought a number of their competitors.

The information age is doing the same thing with data - data becomes more valuable when you have a lot of it. One picture is worthless, but a billion allows you to extract relationships and features. Today we are continually seeing large tech companies acquire smaller ones/

Net effect is we have industries conglomerate until you're left with a few powerful entities who point at each other as competition but ultimately collude.

You need strong regulation and oversight to manage that, and a government supportive of labour. Will that happen in a system where you only have to donate to two parties? Or threaten election integrity by donating to the party that doesn't threaten your power? That may be a tough ask.