Long post ahead, be ye forewarned.
I watched an interesting video recently on the effect of automation, robotics and artificial intelligence has and will have on humanity. You can view the video
When I think about the situation, I find that these changes won't be as sudden as he argues they will be. My thoughts on how this all plays out is the following: For example, going back to automatically driving cars, I do not believe that automatic cars will be allowed to drive by themselves for quite some time after their mass production. Instead, there will have to be a driver there who supervises the car during the trip just in case. Likewise, they'd need to be alert and not under the influence of drugs and alcohol. Just like those who supervise learner drivers are held to the same standards as if they are driving themselves.
Small businesses won't be able to afford robots as they require too much of an upfront cost and the larger ones will experiment by taking a couple on board while keeping a majority of humans (kinda the same process supermarkets are taking with self serve checkouts). A lot of countries will probably at this point see the effect that these displaced jobs will have on the economy and probably introduce regulations on the percentage of robot workers to humans. Human oversight on operations will be required for a long time and thus the last of the jobs to go will probably be managerial and supervisor positions.
Poorer countries will be less likely to use robotics at all and further fall behind, some of the growing economic countries like india or china might be tempted to get into robotics, but this could reverse the progress they've been making in fighting off poverty. Some might take the bite, others probably won't. The 3rd world will probably fall drastically behind again, but catch up quickly once the west figures out how to handle this new dilemma and decides to help build their economies.
So, the government response will initially be regulation; preventing businesses sacking vast majorities of their employees for robots. Not that they'd do this anyway as they'd start off experimenting with a few here and there. They'd still be pretty damn expensive so they wouldn't be gung ho, and those working for small businesses wouldn't have to worry about their jobs for quite a while until the prices for robots dropped dramatically.
The government would require a lot of automated jobs to still be supervised by staff (such as self driving cars) and that at this stage, a long term plan will be developed about what they're going to do about supporting people who will be out of jobs. I do suspect that the short term solution will be some sort of ramped up welfare system where everyone who is out of jobs will be given a minimum wage by the government. This will keep the capitalist system going for maybe quite a long time but it's not viable in the long term.
Obviously the wealthy who are benefiting from this system will perpetuate it for as long as they possibly can but there will come a point where money is essentially worthless and some sort of socialist or communist like system will come into effect mostly because the government will take it upon itself to offer the same services that these wealthy individuals are involved in and will undercut them to keep prices cheap because they'd be required to do it to keep the world going.
The wealthy will eventually be eradicated and the whole concept of paying for things will be demolished. People will take as much as they need, with perhaps some sort of upper limit on what they can take, though this probably won't be needed to worried about as at this point it wouldn't really benefit anyone being greedy. Nations will connect to each other at this point in a even more cohesive version of the EU allowing for free travel between nations and people living basically wherever they want whenever they want.
The only potential problem I see here lies in the government as while all other jobs may go, the government will persevere and as it takes a larger and larger role in supporting the populace, there is an extremely large danger of it turning into a totalitarian state. The shift of this happening or not will be at the later stages and I think people will need to be very careful, and very vigilant to ensure that this doesn't happen.
Lastly, I think this is probably much further away than the video makes it out to be. I get the impression that they think this will happen in the next 5,10 or perhaps even up to 20 years, but I suspect they are going for the former figures rather than the later figures. I however suspect that this will not happen this quickly. I think we will probably be seeing these sort of big changes happening in perhaps 50 years, or maybe even more.
Automated cars for example are something that the author believes we are on the precipice of releasing en masse, but the fact of the matter is there is so much work left to do. Big car manufacturers (i.e. not Google) have stated they will be releasing basic types of these self driving cars in the next 5 or so years, essentially they are like "cruise control plus". A far cry from your own personal robotic chauffeur. The ethical issues that this equipment has to deal with still needs to be fleshed out.
My ball park figure for this? I think maybe, at the earliest it will be 15 years away till we see any sort of large scale deployment of self driving cars and probably 25 years away from robots being involved in easily automated type jobs. I actually think it will be even longer, but I am somewhat hedging on these earlier dates. Now, why do I think that this will be so far away? I'll illustrate with an example.
Let's look at self serve checkouts at a supermarket. The first of these came out in the early 90s, but have only really taken off in the last couple years. I think the ones in the early 90s would have been rather limited, but those from the early 2000s onwards would have the processing power to do what we can do now so I'll take the figure from this point in time rather than the 90s.
Even with this figure in mind, the time from when these machines were first being taken on board to their near ubiquitous presence now has been about 15 years and i most supermarkets you go to there are still plenty of staff working in lanes. It will probably be another 10 years till these lanes are your only option, with some staff on board making sure you don't pinch anything. Thus the full effect of this technology on employment will have taken 25 years. Seeing this, why would we assume that robotic automation will be near instant 100% permeation upon its release?
Thus, we have a long time to respond to the effects these inventions will have on the economy and it is not going to be a panicked rush that the author suggests it will be. Even assuming the implementation takes half the time I suggest, that is still over a decade to respond to the ongoing change, and that is a decade on top of whatever time we have now till their first commercial release, which as I suggested earlier is still over a decade away anyways. So yeah, I think we can cool our heels for a while and not panic. We do need to respond, but we will be given ample time to do so.