First of all, thanks to the OP for starting this discussion, I have to say it's quite engaging; just read through the whole 20 pages, and the arguments on both sides provide some good food for thought.

It's obvious that people are separated on politics and personal value lines, although it's a bit disconcerting to see so many in support of what boils down to essentially communist solutions. After all the pain and suffering caused in the past 100 years by this brilliant on the outside/rotten in its core philosophy one would have hoped to see less enthusiasm for repeating the mistakes of the past...
http://www.thecommentator.com/article/4230/so_how_many_did_communism_kill

Anyway, something that I haven't seen challenged too much so far is the assumption that the people who lose their jobs to automation will only have two choices: starve or be on (various forms of) welfare. I think this is a false dichotomy, but to explain why we may need to take a step back and look at the overall picture that is being considered here.

We are assuming a society many years into the future, where all the menial jobs can be done cheaper and more efficient by automation (I'll call them "robots" for simplicity but of course this may imply computers/machines/etc.). There is a secondary aspect of even complex jobs being taken over by robots (doctors, lawyers and programmers) but that is a non-issue in the discussion, simply because of the very small number (relative to the population) of people engaged in these jobs. If somehow all the doctors will be replaced tomorrow by Watson, they can always fall back to flipping burgers at McD; they won't be happy, but they won't riot either. The problem appears when there are no jobs to fall back on, no McD, no agriculture, no wiping the floors, nothing; that's when you have huge numbers of people with no marketable skills and no possibility for employment, and that can indeed lead to social unrest.

But let us pause for a moment and look at this world that we're imagining -- it can not have just manifested out of nothing, it must have reached that stage through change and adaptation, in other words through evolution. Just like humans did not suddenly appear from amoebas, robots will not suddenly appear one day and instantly "fire" all burger flippers and factory workers. It takes time and a lot of changes to get from the world as it is today to the world we're talking about here. Robots will be extremely expensive at first, cumbersome and limited; for a while a simple reduction in benefits or salaries will be enough to keep humans profitable, even while robots already exist.
http://www.dailytech.com/Foxconn+Runs+into+Trouble+Deploying+Robot+Replacements+for+Human+Workers/article29390.htm

Eventually, a company will appear that employed a 100% robotic workforce, and managed to make products cheaper and of better quality than their competitors. Even at that point, it would still be years before the large existing corporations would adapt their entire production to the new paradigm -- they always suffer from a lot of inertia and risk-aversion, so the improvements offered by the robots will need to be really worth it to the bottom line in order for them to engage in such a move. A robot will not only have to be cheaper than a human, it will have to be significantly cheaper and more efficient.

So what happens when we reach the point of all jobs being automated? Well, for one thing the robots themselves will be very cheap at this point, flexible, resilient and ubiquitous. They would have to be, it's what needs to happen with any technology before it takes over businesses on a global scale. It's what happened to mobile phones (once the price of a small car), computers and networks. Even in the poorest places on Earth, one can get a simple mobile phone and a netbook for next to nothing -- they may not have running water or enough food, but mobile phones are literally everywhere, including remote areas in Africa and Asia. Robots will also have to get there at the time of them "taking over".

But wait -- this means that even those poor, dis-enfranchised people will be able to afford robots for themselves! Remember they have to be significantly cheaper than the humans they replace, yet more efficient. Sure, not the latest model but a refurbished 10-year old one; not a complex life-companion but a much simpler one, still capable of doing a lot of menial work (that's what they've been created for, right?). They will be able to build them from parts, get obsolete models for free at corporate hand-overs, there will a "one robot per family" charity sending them all over the place... Basically even poor households will have access to their own obedient servants, ready to work 24x7 to provide for the family: they can build houses, grow crops, tend to farm animals, dig irrigation canals, fall trees, drive vehicles, care for the young ones and so on and on. A small village may be able to even get a team of 10-20 robots to work for them and build large projects. Again, this will not happen over night -- first one guy will be fired from the factory and buy a robot with his severance pay; then others around might notice the benefits and borrow the robot from time to time for their own jobs; then the second person does the same and so on and so on, by the time everyone lost their jobs robots will already be present in their community.

The end result of all this will be people with no jobs, yet able to survive based on the labor of their robots. In a weird way one may perhaps view this as another take on the "universal income" option above, but one that emerges from society and does not involve state intervention at any point. Actually, the only problems I see with the above scenario would come because of state intervention or human prejudice: the state may impose extremely high import taxes on the otherwise cheap robots, in order to "protect jobs" (ouch!); some "well-intentioned" politicians may be able to ban robot ownership in order to "protect the children"; or people might just hate and destroy them outright out of some irrational Luddite hatred.

Another thing to point out is that human labor will never disappear completely, even in an otherwise fully-automated society. Leaving aside the idea that programming would be done by AI (at that point we will have reached technological singularity, and nothing afterwards could possibly be predicted from our perspective), there will always be a demand for the "human touch". Even if robots would produce demonstrably better products, humans are not entirely rational creatures (thankfully) and as such they will prefer a certain sweater simply because of the "made by human" sticker. Just as we now have organic/fair-trade/hand-made products which actually are part of a growing market, we will also have "human-made" objects still being produced and valued by a significant proportion of the population. Many people will also reject robots out of hand (for whatever reason) so they will choose to employ human baby-sitters, gardeners or secretaries -- it's something that can be seen with racism and sexism in the workplace even now, there is nothing protecting robots from the same treatment.


One thing is clear and undeniable: the world will change, and quite substantially at that. There will be pain and suffering, just as there always has been since the dawn of civilization -- the solution to that however will not be found in world-wide statism, tyranny and communism. I wish we as a race will have enough reason and empathy to make the right choice here -- although, anything is possible (and on-going world events do not provide much hope, unfortunately).