Skip to main content

A Robot Ate My Job!


The World Economic Forum recently over Davos for its annual pow wow. On the agenda were all kinds of delights, including some discussion of technological unemployment down to software and robots displacing human labour due to their increasing cheapness relative to human labour.

Should we all be worried? Well, it depends what line of work you're in. According to a piece on the WEF website those folks doing low-skilled jobs are at greater risk than those doing more-skilled work. Basically it breaks down to; how easily can a machine undertake a task to a customer's satisfaction?

That will be easier with lower-skilled work because it's relative repetitive and usually the boundaries of the work tasks are narrow. That means there is a process with set steps and a very simple workflow for dealing with different possibilities, and that the number of those possibilities is fairly limited. The narrower those bounds, the more easily the task can be automated.

This isn't a discussion of productivity, but rather of the difference in cost to a business of employing humans versus employing robots and software. Costs associated with human labour include payroll and employment taxes like healthcare mandates and social security, while robot and software costs will be the purchase of the goods and/or services that are the robots and software themselves from their vendors plus technical support and maintenance.

There is another cost not elaborated above. Due to incompetence humans or robots may under-perform to a degree sufficient that it puts off even long-time customers and costs the business a great deal of revenue. The trade off between human work and automation is right here. Will customers be sufficiently worse off with an automated system that they go to a competitor that uses humans for the task in question?

This could potentially scupper many automation initiatives as businesses throw money after cost-saving automation only to see their revenues fall faster than their costs. Whoopsy!

Rather bizarrely the piece reads in places like something from Salon or Mother Jones - with mention of automation being only good for the rich - and seems to be intended to alarm considering the 5 million jobs figure which will supposedly be automated away by 2020. The thing is, however, that 5 million shared between the entire OECD plus BRICS is nothing, considering the labour markets of these dozens of countries employ over 1 billion people.

The author does acknowledge an earlier piece from Business Insider that was far less hyperbolic about the likely change to the jobs landscape over the years to 2020;
WEF said in its report, entitled "The Future of Jobs," which was published on Monday, that while skills and jobs displacement will affect every industry and geographical region, these job losses can be offset by employment growth in other areas.
There is no doubt that robots and software will eventually do the vast majority of the important tasks currently undertaken by humans, but time-scales are impossible to confirm because nobody knows how long it will take to write algorithms that can account for the possible combinations of tasks taken in by any given role.

It is likely that, in fact, facets of jobs will be automated away rather than whole roles, and so the number of, say, data entry clerks will decrease but the data entry clerks will continue to exist using fewer people to achieve the same result due to process automation. This will go on and on throughout the 21st Century until nearly everything we do for ourselves today has been automated away.

When? I don't know. Worse, I have no informed idea, only a vague inkling that it'll be somewhere between 2075 and 2200. Enjoy the wait!

Comments

Popular posts from this blog

Private Ownership and the Emergence of Field-based Agriculture

Quick update: There is a nicer, fancier article on this very subject on another blog. If for some reason you read my article below, treat yourself and partake of properal's piece too . ~~~ There is a paper by Samuel Bowles and Jung-Kyoo Choi called 'Coevolution of farming and private property during the early Holocene' and it is wonderful. It leaves a few stones unturned and its thesis needs to be empirically verified or falsified but it really begins to clarify the intimate relationship between the form of agriculture that we refer to as farming on the one hand and private ownership on the other. Their thesis is that technology was not the driver that led to long-term (inter-generational) farming, but also that farming did not follow some moment where the folks in a society all said "hey, let's all have private property now!" Rather, what they posit is that farming and private property actually coalesced, ad-hoc and over a multi-generational time-fram...

I AM AN AUSTRIAN

Is it so wrong? Really? Just humour me, dudes and dudettes. I am an Austrian. I am a Libertarian. I am an Austro-Libertarian. I'm evidently also a hypocrite, as I've used most of these words without capitals in past posts. Oops. I've made Austrian economics my home because it accords better with certain concerns of mine; why have a subjective theory of value and then lump desires and capacities into aggregates? Why declare that economic facts can be gleaned from the movements of particular markets at particular times in the past? Rothbard sums up the problem with both phenomena in a way that no mainstream economist ever would, since to do so would be to admit that there are entire fields of modern economics that are, at best, pointless, and at worst, harmful. NOT MAINSTREAM? Why is Austrian economics not mainstream? It rejects the efficacy of aggregates and mathematical formulae to arrive at economic truths. According to the Austrian worldview,...

ECON 1c: GOOD GOD, IT'S THE GOODS!!

I still reserve the right to be wrong. So, goods. Goody gumdrops. It's a good thing, in a blog about economics, to talk about goods. Hopefully I'll do a good job, because I'm running out of tired turns of phrase. SCARCE AND NON-SCARCE RESOURCES So, there's some stuff you can get as much of as you like ( I can download the pdf of Jeffrey Tucker's "It's a Jetsons World" ) and there's some stuff you can't. The latter would include computers. But let's use an easier example; that pdf is a digitised version of a book, and books in their hardcopy form are scarce, as opposed to the intangible text inside. Air is another non-scarce resource. NON-SCARCE GOODS Non-scarce goods are called goods because they can be consumed by humans. If they have their origin in human action or production they can also be called products (as per ECON 1a). You can consume air, so air is a good. As regards produced goods, if they are infinitely repro...