Looking at the average earnings on these sorts of platforms it would seem like a challenge to eek out even a meagre existence in a first world country.
This is also a much bigger opportunity for startups than traditional eyeballs model. A startup that can pull it off could easily be worth a -trillion- dollars. Apple, Microsoft, Google grew large just by taking $100-200 a year from a billion users. Imagine a mechanical turk that grew to a billion users - it could be taking $500 a month from users while paying them $5000 a month. Truly immense potential.
Higher skilled sellers who can demand higher prices will want to avoid these platform for that reason. They don't want to have the platform owner taking a cut when they are capable of finding customers on their own. This leads to a sort of lemon market where the buyers who are willing to pay the highest rates for the best workers will tend to avoid these places altogether.
So yes, they are commodities, but if they are a lot of them, and a lot of work to go around, wages rise. We saw this with the industrial revolution, especially assembly lines. Mid 20th century you could live pretty well as an unskilled assembly line worker, but then the work dried up.
Most modern work, knowledge work, hasn't been turned into assembly line yet. Many would argue that's because it can't be, but I disagree with that, much of enterprise software is about turning it into that (workflows,processes). It just requires more of the same. There are actually very few cloud labor startups (mturk, crowdflower, mobilworks) with at most a 100 programmers, one reason being that it doesn't have the network effects moat that VC's look for. Robotics is another casualty of that. How do you defend your cloud labor,robotics company from competition? That's why it will have to be built from the ground up by bootstrappers.
It also makes the competition for knowledge work go international which might increase pay in developing countries but lower it in first world countries.
Higher pay rates exist for two reasons, unfair monopoly positions (e.g. bill gates and the OS monopoly of the 90s, actors/athletes), and because there is complex work that hasn't been deskilled yet. Not much can be done about the former except regulations, the latter is where programmers can change the situation. Demand is everywhere - whether it's fulfilled by cloud labor, or by specialized labor - is the question.
Either way, I think we can agree that immense opportunities exist for startups to profit from this space.
If you equalize wages you also need to equalize the cost of living which is something that doesn't seem to be happening because a lot of people have a lot of money locked up in land etc that they are not going to want to see devalued.
If deskilled work does not cover rent payments then it is not much good to anyone.
Yes, you're totally right - land will get devalued. The older generations in richer countries will lose the most relatively. This is not a bad thing fundamentally (it reverses the rich get richer trend), and is actually what's already happening, but people don't feel it. Every story that is heard about poverty reduction in China is also a story of westerners getting relatively poorer - though in absolute terms everyone is getting better quality of life.
I'm not sure that the older/richer will lose out. They have a lot of political clout that they will bring to bear to keep value in their homes and retirement assets. A lot of this will be hoarded rather than redistributed so demand will continue to outstrip supply in many areas.
Deskilling benefits everyone, is leads to more equitable wealth distribution. The people inventing the assembly lines (hackers) will be paid more than the assembly line worker (cloud laborers), but that's a fair reward for the effort. They won't be much richer (just like mechanical engineers who invent factory equipment) because fierce competition will prevent it (there aren't network/platform effects that lead to gates/zuckerburg).