Here is Sam Harris, in an addendum to his post How Rich is Too Rich?

I would be interested to know if any economist has an economic argument against the following ideas:

Future breakthroughs in technology (e.g. robotics, nanotech) could eliminate millions of jobs very quickly, creating a serious problem of unemployment.

I am not suggesting that this is likely in the near term…. I am suggesting, however, that there is nothing that rules out the possibility of vastly more powerful technologies creating a net loss of available jobs and concentrating wealth to an unprecedented degree.

and here is Mark J. Perry’s response

Yes, there IS a fundamental principle of economics that rules out a serious persistent problem of unemployment:

The first principle of economics is that we live in a world of scarcity, and the second principle of economics is that individuals have unlimited wants and desires.

Therefore, the second principle of economics: unlimited wants and desires, rules out any long-term problems of unemployment.

Am I the only one that can’t fill in the gap between those laws of economics and the long-term problems of unemployment ?

To be sure, history of technology supports the optimist. As Perry mentions in his blog, the industrial revolution and our own information revolution have openned new jobs as they were killing old ones. Still, History’s laboratory only provides examples in which machines are not sophisticated enough to compete with us monkeys on everything. Even in low-skilled jobs like driving, cleaning, cooking, we are still doing better. Perhaps this will always be the case, but the laws of economics are not in contrast with the possibility that machines become so sophisticated that they outperform us, or at least a large chunk of us, in everything. In such a situation, with unskilled human labor having a cheap substitute, natural resources might become the binding constraint for production.

Here is an historical analogue of the situation I have in mind: The Roman empire suffered from a type of substitution for human labor that our Silicon Valley have not yet been able to come up with: slavery. And the story of the empire matches Sam Harris horror forecast, with wealth (that is to say, land) concentrated in the hands of very few, and a persistent unemployment in cities crowded with unskilled  citizen who are unable to compete with slaves in the jobmarket.

Now, I suppose one can construct a model, and maybe this is what Perry refers to, that shows that in an ideal world without a welfare state people should still find jobs as long as there is some epsilon value in human labor, with a salary that reflects this value. Indeed, even the unemployed plebs of Rome have managed to scratch some sesterces by provididng whatever utility the few rich aristocrats got from having flatterers and clients welcoming them in the morning, and one can always work as a thief or a beggar. But overall, I think it is fair to say that the benefits that slavery brought to the Roman jobmarket is closer to Harris’ prediction — net loss of available jobs and concentration of wealth — than to what the industrial revolution brought to western civ.