This is pretty standard economic theory. We'd call automation "productivity gains". If tax rates are structured correctly, we expect productivity gains to be redistributed to workers - as a worker can produce more, their income rises. What we've seen in the past 30 years is that instead of productivity gains being shared with workers, they've been systematically redistributed upwards.
It's not a problem a machine can replace 10 workers. You'll still need someone to manufacture and maintain that machine, and those people will be paid more than the workers that were replaced. The other workers will go on to other jobs.
Per Dean Baker[0]:
> Replacing human labor with technology is a very old story. It’s called “productivity growth.” We’ve been seeing it pretty much as long as we have had a capitalist economy. In fact, this is what allows for sustained improvements in living standards. If we had not seen massive productivity growth in agriculture, then the bulk of the country would still be working on farms, otherwise we would be going hungry.
> Replacing human labor with technology is a very old story. It’s called “productivity growth.”
This is true and indisputable in the trivial sense, in that it describes history accurately. However the problem I have with this sort of statement is that it implicitly assumes that the future will be the same as the past. In this case I believe there are reasonable arguments that the future might be different.
To expand upon that point, the position you've laid out says nothing about the rate of destruction of jobs. It seems plausible (though obviously not certain) that the next decade or two of automation could involve rates of obsolescence that have not been seen before in history.
If "productivity growth" happens on the timescale of a generation, then the next generation can learn the new skills, and there is manageable social impact. The historical examples I'm familiar with fall into this category (e.g. power loom adoption over 25-50 years in the first half of the 19th century). In this domain, the rate of job destruction is below the maximum rate of job creation, and so there is no net job destruction.
However if that effect happens in a shorter timeframe, then you could have a large segment of the population all trying to find new work at the same time. (e.g. a self-driving truck makes most long-distance freight drivers obsolete in 5 years, removing 1% of jobs from the economy without an immediate replacement; in the worst case, rinse and repeat for 10 other industries in the same period). In this domain, the rate of job creation hits a limit, and therefore there is net job destruction, until the rate of job destruction decreases and job creation can catch up.
Do you have a model for where this breakdown might occur?
> The other workers will go on to other jobs.
This is true _on a long enough time scale_, but economics is silent on the broader societal and political impact during the re-equilibration period following a discontinuity in employment like this.
It's not a problem a machine can replace 10 workers. You'll still need someone to manufacture and maintain that machine, and those people will be paid more than the workers that were replaced. The other workers will go on to other jobs.
Per Dean Baker[0]:
> Replacing human labor with technology is a very old story. It’s called “productivity growth.” We’ve been seeing it pretty much as long as we have had a capitalist economy. In fact, this is what allows for sustained improvements in living standards. If we had not seen massive productivity growth in agriculture, then the bulk of the country would still be working on farms, otherwise we would be going hungry.
[0]:https://www.patreon.com/posts/robots-taking-24450257