The dynamics of automation cannot be properly discussed without considering who controls the means of production.
Who owns the robots, and why does this matter? To understand why the questions of ownership and control are important to understanding the future of automation, I think it is useful to look at the rather amusing case of a tech worker who, in early 2013, was found out to have been outsourcing his own job to China. From BBC:
A security check on a US company has reportedly revealed one of its staff was outsourcing his work to China.
The software developer, in his 40s, is thought to have spent his workdays surfing the web, watching cat videos on YouTube and browsing Reddit and eBay. He reportedly paid just a fifth of his six-figure salary to a company based in Shenyang to do his job.
Unsurprisingly, the worker appears to have been fired.
It’s interesting to contrast this particular case of outsourcing with what is generally thought of in typical discussions. In general, outsourcing is something that draws the ire of leftists, progressives, and labor activists in the United States. One of the primary arguments (if not the primary argument) against outsourcing is how it undermines the position of the American worker, and how it has contributed to stagnant wages and standards of living for the past few decades.
And yet, in this particular case of a worker taking it upon himself to outsource his job, it is clear that the worker has benefited greatly from the arrangement; he pays $50k/year to a Chinese consulting company to do all of his work, for which he receives ~$250k/year, and then spends the whole day lounging about and watching cats do funny things. (For the record, I think this video of a baby bear and a baby wolf playing trumps most cat videos).
(Also, the fact that the worker in question is making a salary that puts him rather close to being part of “The 1%” should probably be shelved for a later discussion; for now, lets just go with the standard Marxist definition of class, in terms of how much control an individual has over the means of production: factories, mines, breweries, power plants, and so on).
The difference between “bad outsourcing” (“bad” from the labor/leftist perspective) and this case of outsourcing should be obvious. “Bad outsourcing” is done on the company’s terms–that is, in a situation where the worker has no say in the matter, and where the company boss outsources a job and pockets the difference between cheap foreign labor and American labor. Outsourcing by the workers, on the other hand, is just that: outsourcing done on terms dictated by the worker, where the worker outsources her own job and pockets the difference.
Now, extend this observation to automation. As it stands, there is a whole lot of concern (particularly, it seems, from economists) that the increasing rate of automation, roboticization, and cybernation is creating a secular decline in employment, leaving an increasing number of millions structurally unemployed, and severely limited in their ability to access the theoretical benefits of mass automation. But just as with outsourcing, the underlying reason why workers are losing out is because they hold little to no power in the process of implementing automation. Labor does not control the means of (automated) production; capital does.
When a company decides to automate their factory, the workers are simply replaced–they gain nothing, and lose a whole lot. The cheaper marginal cost of automated production is then reflected in the profit margins of the company and in the reduced cost to consumers (a group which may or may not overlap with the newly unemployed workers). But imagine a scenario where instead of replacing the workers, automation was implemented by the workers, in a manner similar to how outsourcing was implemented by the tech worker discussed above. Perhaps the workers decide to make use of their free time by teaming up, taking some courses on mechanical design and feedback control logic, developing an unnecessarily complicated QA procedure, and automating the factory themselves. Now, the reduced costs from automated production are reflected in far more leisure time for the workers and in reduced costs to consumers; and its debatable whether the company even loses out (putting aside questions of opportunity costs), since the rate of profit remains the same.
This thought experiment emphasizes the importance of understanding who holds power in a capitalist economy, and what this implies for how automation, roboticization, and cybernation will affect wages, employment, and living standards. If workers, and the masses in general, are not in control of how technology is implemented, then the benefits will be disproportionately (and in some cases, entirely) accrued by the already wealthy and powerful. On the other hand, if common people can take the initiative to study and learn the technical and scientific knowledge necessary to take control of this process, then the benefits will be spread across society in a much more free, open, and equitable manner.
This is why I think movements like the open-source movement, and institutions like Copyleft and Creative Commons, have so much potential for radically altering our political economy. These movements help ensure that there is at least some level of free and open access to the educational and developmental resources necessary for the general masses to engage in autonomous research, development, and implementation of high technology, in ways that socialize the benefits.
The main limitation of these organizations and spaces, however, is that they seem to be mostly populated by people who are otherwise already engaged in science and technology, and are pursuing free and autonomous development as a hobby. The people who would most benefit from learning and developing their own robotic or cybernetic systems are those who are absent from these spaces: low-skilled workers, unemployed people, marginalized inner-city communities, and so on. Thus, there is a huge need for people with a technical background to transfer their skills and expertise to people without such skills, and to help establish cooperatives, non-profits, organizations, etc. that enable workers and communities to develop and implement high technology on their own terms.
The potential for such a movement is huge. Both types of spaces–community/worker organizations, and open-source, pseudo-anti-capitalist technology organizations–exist. The immediate task, then, is to establish networks between the two spaces, and create and expand projects that re-appropriate technology for the good of general society. This would place us on a path to putting the means of production and distribution into the hands of communities and workers–and ensuring a future where automation will benefit all, rather than the few.