How and why today’s companies boost productivity by using technology to tap a vast reserve of willing, free labor – you.
I’ll never forget the first time I got fired.
I was a kid in Portland, Maine, working as a gas station attendant. (Remember gas station attendants? Probably not, unless you’re over 35.)
Anyway, after a long and arduous ten-hour shift standing in the rain and sucking fumes, I balanced my “cash drawer” (which was in reality the pockets of my jeans) and came up 50 cents short. Fifty cents. I had pumped nearly 2,000 gallons of gas that day, but that didn’t matter. In the eyes of my boss, fifty cents short was fifty cents too much.
The rule at this place was simple – if your cash didn’t balance out, you were gone, period. The company who owned this station ran several dozen other stations just like it, and I guess they figured that two bits here and two bits there might eventually, somehow, add up to real money.
It was a few years later that this company installed self-serve automated pumps in all of their stations, and fired every attendant in their employ. No more attendant wages. No more of the associated worker’s comp premiums, unemployment premiums, or matching FICA payments.
Productivity swelled. The company literally saved millions, which they eagerly pocketed as profit. And guess what? They never came up short on a shift again, either.
Did these automated pumps eliminate the actual task of getting the gasoline out of the ground and into the cars? Of course not. But now, the station was able to demand that customers do that. Oh sure, some were reluctant at first. But as soon as they realized that every station in town was jumping on the self-serve bandwagon (hey, gas station owners are no dummies), the options were pretty clear – open your own petroleum refinery, take the bus, or learn to deal with the indignity of being coerced into service as a lowly part-time pump jockey. Suck fumes or walk, pal.
Little did we know that self-service gasoline was only the tip of the iceberg.
The first automated pumps were expensive and s – l – o - w, communicating at 300 baud over regular lines leased from the local telephone company. But like all technology, pumps became much cheaper, much faster, and much more capable. Running a station became more profitable than ever. More and more stations opened, and larger stations soon became the norm.
The Great Productivity Myth
It was not long before other industries sat up and took notice.
Today’s businesses aren’t saddled with the burden of massively expensive hardware and dog-slow data transmission. PCs are practically disposable, servers are dirt cheap, and bandwidth is now a commodity. This revolution in hardware, combined with the big ol’ Inter-tubes and the technologies that run on it – the World Wide Web, email, and the rest – is allowing enterprises of all kinds, from the airlines to your local grocery store, do what the gas stations did.
They too have enjoyed unprecedented gains in productivity. If you asked their accountants how, they’d tell you they did it the old-fashioned way – by reducing costs. From an financial point of view, that's technically correct. But from a societal point of view, true costs are not being reduced at all, but merely shifted off the books – a macroeconomic sleight-of-hand that places inputs outside the purview of the enterprise.
In any capitalistic endeavor, the most expensive input by far is labor. Here in the US and elsewhere throughout the developed world, the cost of labor per dollar’s worth of output is dropping precipitously. Is this because there’s less work to be done? Or is it because true labor costs aren’t being accounted for?
Think of it this way – if I pay a high-school kid to put my gas into your car, that’s a cost to me. If I get you to do it, it’s not. The gas is still getting into the car, but this way, nowhere on my books do I have to account for that fact. See how that works?
Artificial Artificial Intelligence
With each passing year, American consumers are becoming increasingly pressed into a form of consumption-based indentured servitude, made possible by cheap hardware, cheap connectivity, and cheap, ubiquitous Internet access, and businesses are reaping the benefits.
Twenty years ago, if you had an issue with your bank account, you’d ask a teller. Today, you are instructed on how to find that information yourself; in fact, many banks now charge a fee to even speak to a human teller. Fifteen years ago, most people wouldn’t have dreamed of booking their own airline flight. Today, not only are we expected to book our own travel as a matter of course, we print our own itineraries and our own boarding passes, and check our own bags, too. Grocery stores with manual checkouts are served by one casher per terminal. In those now installing self-serve checkouts, a cashier handles six terminals. May God help you if the sticker falls off your banana.
In the tech world, not all that long ago, it was just assumed that knowledgeable, competent technical support was never more than a phone call away. Today, if you are fortunate, the company you’re dealing with may offer a support “package” – one that requires you to pay for the privilege of speaking to a human, and pay even more if you want to assure that human speaks the same language as you. Free technical support has now been pushed entirely online, to be found in the form of incomprehensible wikis, outdated FAQs, esoteric so-called “knowledge” bases, or online forums where customers are not only required to help themselves, but are encouraged to devote their spare time helping each other, perhaps predicated on the belief that when frustrated users find themselves in the same boat, they will attempt to row together.
The Web has turned us into our own bankers, our own travel agents, and our own support desks. It has turned investors into their own brokers, students into their own teachers, and patients into their own doctors. While many of us continue to perceive the Web as a communications/entertainment/social networking medium, many businesses clearly see it as something else – a means of leveraging a motivated, compliant, highly scalable workforce on-demand; in effect, a giant Mechanical Turk.
I’m certainly not saying that this is all bad. Like most people, I enjoy the convenience of being able to check my bank balance online, for example. That is what the Web is good at – pushing data around. That is, in fact, the only thing the Web is good at. If you can take a problem and accurately express in terms of a simple data transaction, then that’s a good Web application.
The problem lies in how organizations are attempting to shove increasingly complex domains into this simple exchange-of-data paradigm. Consider, for example, the fast-growing trend among colleges and universities toward self-service education – supplanting the traditional classroom lecture with online “instruction.” Standardized, cookie-cutter lesson plans allow the use of lower-paid, less qualified faculty. More students can be registered per session. There are fewer facilities to maintain, and no classroom to heat or light. The school then uses the “convenience factor” to justify exacting an additional “technology fee.”
In this scenario, the university wins, but the students lose. That's because online instruction is an imperfect, incomplete model of the real-time, physical, face-to-face interaction that normally takes place between students and teachers, and among students themselves. Forums, chat rooms, and email are merely tools used to emulate the organic nature of a real classroom. In this case, the exchange of data simply cannot truly replicate the exchange of knowledge, or of ideas, or of service; but can only serve as a dim digital approximation thereof.
Pay No Attention to the Man Behind the Curtain
In a world where institutions of all kinds have long ago squeezed every penny out of the supply chain, those institutions are being compelled to automate more and more of their processes in order to boost productivity, which brings us to the real issue – the inexact and imperfect art of modeling complex “meat-world” business processes and transactions in software. Since all software can do is manipulate data, the success of a given solution lies in how well it models its problem domain as data.
When it comes to checking my bank balance or transferring funds, the Web is great. That’s a pretty simple transaction; one that can be expressed perfectly in terms of ones and zeros.
However, with more complex processes such as educating our students, as we’ve just seen, technological models often fall short.
And in those frequent situations when a piece of the domain puzzle can’t be modeled inside the system at all (think gas pumps here), it then falls upon the actor in that particular use case (read: you) to provide the missing “functionality.”
Futurists once told us that “smart machines” would one day open up a utopian world of decadent leisure, but that's not happening. Let's not kid ourselves. We’ve seen the future, in which the old dreams of artificial intelligence are being replaced by a new reality in which “intelligent” systems use thier own users as actual components in the application.
Welcome ladies and gentlemen, to the super-duper, ultra-convenient, self, self, self serve world. It appears we have truly found the ghost in the machine, and it is us.