Wednesday, July 15, 2015


Thoughts on “Rise of the Robots”

In what feels like a previous life, I used to read a lot of midcentury social theory.  Some very smart people -- I’m thinking here of folks like David Riesman -- used to argue that the great crisis of the coming decades would be the sudden abundance of leisure.  Given the outsize productivity gains of the previous decades, and the declines in the average workweek, it seemed reasonable to look forward and suggest that the future held ever-shorter workweeks.  In that scenario, what to do with so much “found time” seemed like a burning issue.  (Riesman titled one of his books Abundance for What?, which gives a pretty accurate idea of what he took as given.)  No less an economic thinker than Keynes believed that workweeks would get steadily shorter, freeing people up to pursue more interesting things.  

Somewhere along the way, that didn’t happen.  

Productivity is much higher now than it was then, though living standards for middle-class Americans peaked somewhere around 1970.  We’ve become astoundingly good at coming up with ways to consume leisure time.  My kids have trouble believing that there were once only four channels on tv, and the idea that we didn’t have the internet strikes them as horrifying.   I’m old enough to remember when seeing a movie meant hoping that it came to the local theater, or that it would be “movie of the week” on one of the four channels.  Now, I have to explain to the kids why they shouldn’t stream Netflix over 4G.  The “abundance of leisure” problem has been solved.

But the “how to make a living” problem is real.  Rise of the Robots, by Martin Ford, makes a strong case that technology-driven job elimination is outpacing technology-driven job creation, and it isn’t confined to blue-collar jobs anymore.  As robotics and artificial intelligence have advanced with increasing speed, they’ve been able to displace humans in progressively higher-skilled occupations.  Over time, human labor is split: a small elite either owns or designs the machines, and makes an absolute killing.  A majority is shunted into work that’s both idiosyncratic -- and therefore hard to automate -- and low-value, making it not worth automating. (Much low-end service sector work fits this description.)  And the rest are pushed outside of formal employment altogether.  

In the world predicted at midcentury, this wouldn’t have been a problem.  Many critics at that time assumed that increased wealth would be spread evenly across society; they assumed a de facto decoupling of ownership from profits.  Instead, we’ve doubled down on the tight connection between ownership and profits.  In many fields, we’ve increasingly decoupled work from meaningful income.

Ford notes that healthcare and education have been relatively immune to the rise of the robots so far, with a few exceptions.  (He claims that pharmacists have taken it on the chin; I don’t know whether he’s right about that.)  By the logic of Baumol’s cost disease -- which he briefly outlines but never names -- we should expect cost increases in healthcare and education to continue to outstrip most other sectors, simply because their productivity is increasing much more slowly than everyone else’s.  As that happens, we should expect the pincer movement of higher prices and internal austerity to tighten.  (Ford may be more right than he knows; for example, he places far greater faith in robo-graders for papers than I do.  If I’m right, then the productivity nut is even harder to crack than he thinks it is.)

What Ford gets right, and the midcentury critics got wrong, is that a strong middle class is neither inevitable nor natural.  It’s a relatively recent, and fragile, development, and it was the result of a set of conscious political choices.  As those choices are reversed, the conditions under which a middle class can thrive go away, and the middle class struggles to reproduce itself.  The savvier members engage in “opportunity hoarding,” or pulling up the drawbridge behind them; the less savvy ones gradually (or quickly) lose ground, and wonder just what the hell happened.  

From the perspective of someone working in the community college world, Ford’s diagnosis is bracing.  Community colleges are designed to create a middle class for a society that’s increasingly moving away from a middle class economy.  The task is getting objectively harder.  Jobs that once seemed like sure bets for long-term economic security come under attack, one after the other.  Yes, some still exist, and I’m happy to prepare students for them.  But if you compare, say, the number of people who work at Amazon to the number of people who used to work at Borders, you can see the problem.  

Perversely, I actually draw some hope from the staggering wrongness of Riesman’s and Keynes’ predictions.  They were very smart people who wrote books that captured important truths about their times, but they got the future badly wrong.  As smart as they were, they couldn’t capture the entire picture, and the cracks in their systems let in so much water that the whole thing sunk.  If we’re lucky, fifty years from now, someone will say the same about Martin Ford.  

I take issue with the assumption that a shorter work week was the result of productivity increases. The five-day work week (as opposed to six or seven days) was borne out of a political movement by organized labor, as was the middle class.

Sadly, organized labor and the middle class in general are losing power at a steep face.
I thought Ford's book was brilliant. As someone who used to teach English composition courses at a community college -- thirty years ago, we taught 12 sections a year, 5/5/2, with classes of 27-30 -- and left before I went insane, it occurred to me that the jobs of the adjuncts, teaching assistants, and full-time contingent faculty who are currently teaching the vast majority of all lower-level writing courses could be made considerably easier by getting robots (software) to grade the overwhelming number of papers these people must cope with for their low salaries.

After reading Ford's book, I am certain that robots could do at least as good a job, and probably a better job, than nearly all the English composition teachers, full-time and part-time, at your college. While they should eventually replace them after the full-timers retire, leaving the adjuncts to other work (if they can find it), community colleges should start using robots to grade first-year students' essays as an aid to instructors and a bridge to getting rid of the instructors.

(Now, before I press "Publish Your Comment," I am being asked to put an "X" mark next to the statement "I'm not a robot." I'll lie and do it.)
I keep telling you, you need to read "Player Piano" by Vonnegut. The exact same ideas were expressed in a 1952 novel. His was a mid-century critique that did not include any middle class.

Perhaps a bit from the intro on the Wiki page will help you see its relevance: "The story takes place in a near-future society that is almost totally mechanized, eliminating the need for human laborers. This widespread mechanization creates conflict between the wealthy upper class—the engineers and managers who keep society running—and the lower class, whose skills and purpose in society have been replaced by machines."

From -

The only missing context for you (explaining why 2022 is the near future of a book from 1952) is something more important than Baumol: Moore's Law. That is what tells you that your phone, which is already a supercomputer, will be able to grade papers better than you in a few years because it will rival the power of IBM's Watson. Remember, it was 4 years ago -- two doubling times -- that Watson was on Jeopardy. It is now an MD and could easily replace every pundit on TV if you wanted it to.
Big employers don't want productivity. I mean, they'll take it, but it's not the first thing. The first thing is to win the Class War. The second thing is to make a profit.

Little employers may want productivity, but they're not entirely rational, so . . . yeah.

Mondragon, man. Mondragon.

The prospect of robots, technology, and automation replacing a lot of jobs now considered as being middle-class is an intriguing prospect. Initially, automation began to replace a lot of factory jobs, and fewer people were needed to manufacture the goods we all use. Professional people and the people who were considered as being solidly middle class were initially not affected very much by the onset of automation. After all, it is hard to imagine doctors, lawyers, engineers, or teachers being automated out of existence. But automation is now beginning to replace jobs even in some of these fields. Even truck drivers and taxicab drivers may be automated out of existence by self-driving cars such as those being prototyped by Google. Even the military will become increasingly automated in the future—the Air Force has announced that the F-35 will be the last manned fighter aircraft that they will buy. Perhaps airline pilots may in the future be replaced by sophisticated software which will fly our airplanes for us.

This problem is especially acute in academe. Software packages are now beginning to replace a lot of face-to-face classroom teaching. At Proprietary Art School where I used to teach, the math curriculum has almost entirely been replaced by Pearson’s software. Just about every college course can in principle be moved online, resulting in a need for fewer and fewer classroom teachers. Only research superstars at R1 universities and particularly charismatic teachers at SLACS will survive in the future, with the few surviving faculty of lesser ability being reduced to the status of teaching assistants watching over students hovering over computer terminals.

In the not too distant future, we may see even more stratification into a two-class system, with just a few people who manage and create all of this automation doing quite well and prospering, with the rest of us doomed to descend into the underclass, doing menial low-paying work such as flipping burgers at Wendy’s or bagging groceries at the neighborhood supermarket.

Machines can flip burgers easily. And I bagged my own groceries at the busy self-service checkouts yesterday.
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?