Thursday, January 5, 2012

Recent Reading On Technological Unemployment

In the last month I read two Kindle books about the same topic: How our economy, specifically employment, is going to change in the face on increasing computer automation. Due to automation’s potential negative effect on employability, this is termed ‘technological unemployment’. The books were both excellent and flawed, each in their own way. For the purposes of this review, I think it is important to look at them simultaneously to better assess the books and to help get a ‘big picture’ view of how exactly these concerns apply to ‘the average worker’ (ie – me and you).

The books in question are

“Race Against The Machine: How the Digital Revolution is Accelerating Innovation, Driving Productivity, and Irreversibly Transforming Employment and the Economy”
“The Lights in the Tunnel: Automation, Accelerating Technology and the Economy of the Future"
The links are to the Kindle editions as that is what I read.

Both books make separate but complimentary cases for the potential negative impact of automation of the employability of the population at the current and, following trends, expected future state of things. They both succeed in making these arguments but do so to differing degrees. “Lights In The Tunnel” makes an excellent analogy of, well, lights in a tunnel that I think is one of the major accomplishments of the book. Ford should be commended for creating such an evocative and educational allegory of the modern income & consumer-driven economy. Not only does it demonstrate exactly why an economy with large unemployment is a long-term bad thing on economic grounds, he provides a great easy-to-comprehend tool for people to think about macroeconomics and how to relate big-to-small (ie – them). This is no small accomplishment.

In explaining the coming problem, “Race Against The Machine” is no less convincing but does not provide that an over-arching mental model through which to view all the potential changes that the book is attempting to make people aware of. However, their approach is far more even-handed. In “Lights…Tunnel” Ford’s tone veers towards the panicky. Not explicitly so but his attitude towards automation and view towards where it will all end up both reflect concern without much counterbalance. “Race…Machine” goes to some trouble to point out that automation’s benefits and technology’s benefits in general will act as a counter-point to some of these potential negative effects. This differing attitude in framing the problem has huge consequences when it comes to arriving at prescriptions for handling technological unemployment. I’ll talk about this in a little bit.

The macroeconomic effects of income/revenue dependence in the modern economy is pretty much the punch-line of most concerns about automation. Our consumer economy works because people get paid and use that money to buy things, profits from this are reinvested by businesses in new products and we get new innovations. This is, of course, greatly simplified but this chain of causality is important because if 1/3 of people do not earn money anymore and, in fact, are not capable of earning money ever again then it isn’t just unfortunate for them. Whole industries will cease to exist because the money is not there any more. In addition, our government's financing problems would also become infinitely worse as aside from credit (which is in the process of imploding on a fairly epic scale) the government’s main financing comes from redistributing revenue in various ways. You take that money out and how will any government services continue to work?

I actually think this is a huge, huge deal because all this talk about the Singularity requires something that pro-singularity people do not often deal with directly: In order for all these technologies to be developed a staggering about of money has to be invested in developing them. This investment can and likely will occur in a variety of ways (public/private, standard-venture-capital/kickstarter, etc) but if there is no more income in the economy then it cannot occur at all. Not every aspect of the Singularity will require it but the really, really "Rapture of the Nerds" type stuff definitely does. The other braking effect comes from fact that technologies don’t really ‘mature’ until you make 20million of something on four continents and that most of these technologies are networking-intense so their exponential power really doesn't kick in until you have a lot of them everywhere.

This is how mobile tech is today. We are getting really, really good at incorporating increasing functionality and cost-reduction into mobile devices. This maturity is being driven by the fact that everyone has phones and is willing to pay for financing those innovations. If 1/3 of people can’t pay their cell phone anymore, it doesn’t necessarily stop this innovation but it certainly would slow it down. In the case of technological unemployment, Singularity-type technologies may indeed succeed in developing in the lab but the money isn’t there to put them everywhere or to commoditize them to the extent they become ubiquitous and some of the positive networking effects are seen. What I find really strange is that in both books, even when they mention the Singularity, there's minimal discussion of these potential connections except in a very general nod to "future tech needs research today".

Another more explicable failing of both books is in regards to how they present technological unemployment’s effects on educated versus uneducated people. “Lights In The Tunnel” does better job of pointing out how education does not necessarily protect you from technological unemployment. Ford outlines a great comparison of a housemaid and a radiologist showing that, of the two, the radiologist is far, far more replaceable. “Race Against The Machine” points out the negative effects to the uneducated in quite overwrought terms but doesn’t spend a much time dealing with the fact that certain educated, professional jobs are equally in danger. Where both books fail rather spectacularly is in the prescriptions. Both treat education as if it were magic. If only enough people were educated, then they would be less replaceable. Neither really bother to prove or justify this assertion. For “Race Against The Machine” it’s more forgivable but “Lights In The Tunnel” actively points out how education will not necessarily be protection and then Ford proscribes more of it. It is a pretty dissonant step that decreased the effectiveness of Ford’s argument in general (to me anyway, obviously YMMV).

As stated earlier, the differing frames for the problem especially in light of how “Race…Machine” actively acknowledges the potential positives of Singularity technologies, greatly affect the prescriptions offered in the two works. Ford, while far superior at presenting the problem in an evocative and specific manner, takes quite a negative perspective. We can’t stop technological unemployment only therefore our only option is to delay it long enough to forcibly "freeze" our economy into an income-based one. Basically, use income redistribution to mimic the existing system but without employment. This involves the usual suspects of income redistribution and state control a la Platonic guidance and goodness. It also struck me as completely infeasible to implement and demonstrably unworkable in practice. While “Lights” predates “Race…Machine” and therefore its more simplistic analysis is perhaps excusable for that reason, “Lights” single biggest failing is in its prescriptions. They are so unimaginative, reflexive and ultimately, about preserving the status quo at all costs in the face of immeasurable dislocations that it resulted, in my case at least, in a real let-down of an ending.

Perhaps Ford wished to reduce the amount of ‘new’ or ‘wild’ ideas in the book. But, I think that no matter the historical success of transaction/income based economies it is completely fair to ask "In light of new technological possibilities, do non-income economies become more feasible or more possible and what are the possibilities there?". Both books actually fail to really think about this problem. Ford because I think he was more focused on promoting awareness of this problem without perhaps following through in thinking on solutions. “Race..Machine” acknowledges the unpredictability and complexity of divergent economies via implied associations with non-commercial developments such as Wikipedia. They also do a good job of justifying entrepreneurship on the grounds that the best way to try lots of things and see which are useful is to make it easy for innovators to innovate. But, because this is a scholarly audience and people are generally skeptical about questioning historically proven ideas, they do not really question the income/transaction based model either. I don’t really think we need to tear down the present system (it is in many ways quite fantastic). However, in light of potential game-changers, re-examining key assumptions could be quite useful.

The redistribution of incomes also ignores one consequence of Singularity developments and that is the rise of the “Sovereign Individual”. Right now our redistribution system in the US is highly dependent on star-performers. It works because we have a lot of successful people. But, these are also likely some of the first people to effectively become "Sovereign Individuals" and in doing so, become less bound by state control as it were. This is a huge wild-card factor. Any system proposed to "fix" technological unemployment via redistribution or some state control scheme to forcibly maintain the income-transaction-taxation system will have to intelligently address the new capabilities of star-performers. I'm not just referring to malicious evasions either. Star-performers could innovate services traditionally performed by government. Think the current troubles of the U.S. Post Office magnified by several orders of magnitude.

I will acknowledge that my own libertarian inclinations undoubtedly affect my opinions of Ford’s plan. However, to me, it really seemed that after spending a couple chapters outlining the wild, new and counter-intuitive aspects of the problem of technological unemployment, Ford just phoned-in the prescriptions.

“Race Against The Machine” by comparison, manages to acknowledge the inevitability of at least parts of the Singularity but take a far more innovative and optimistic stance. They actively point out how automation is not a fix-all. They deal with what aspects of our jobs aren’t currently threatened but might be and what involves significant innovations whose solutions are not apparent at present. The authors of “Race Against The Machine” demonstrate significantly more thoroughness in their recommendations because they are actively attempting to make these changes a net positive and do so in a manner that gives far more credence to how people actually behave, what is likely to become public policy and potential interactions between non-economic or non-technological factors for which they cannot properly account. The only prescription which fails in these respects is the education one which I already discussed.

Overall, I would say both works are thought-provoking enough, despite their flaws that they are definitely worth their respective prices. In fact, since reading them, these works have prompted a large degree of consideration on this idea of a ‘Singularity’, the possibilities involved, my own concerns and how people can get a handle on these potentialities. These thoughts are beyond the scope of this review but I am intrigued, curious and eager to learn more. Therefore, in terms of promoting thought and discussion on technological unemployment and the Singularity, both books more than succeed.

Full post here.

Link to full post.