Image: France in 2000 year (XXI century). Future school. From a postcard by Villemard, 1910.
I recently finished listening to the journalist Alexis Madrigal's eight-part podcast series on containers — the large cargo boxes that have revolutionized both international trade in general and the shipping industry in particular. The last episode includes a section on the possible automation of trucking, the last step in delivering so many manufactured goods to the stores that will eventually sell them.
My heart went out to the workers who suffered when containers made their jobs obsolete, but listening to Madrigal's podcast made me wonder about the future of my own profession, too.
According to Madrigal and his guest experts, it is relatively easy to automate the task of shipping goods from one place to another, but the job of trucker includes much more than just driving cargo between ports to big-box stores. Robots can drive trucks, it seems, but they can’t tie down the load or fix the truck if something goes wrong. That's why the task of driving a truck can be automated, but not the job of trucker. That’s a very important, widely applicable distinction.
Like self-driving cars, self-driving trucks aren't quite a thing of the present, Madrigal explains, because of the possibility of so-called “edge cases.” An automated truck can handle an uneventful delivery between a port city and a Target store, but what happens if something weird happens along the way? To use what Madrigal described as a famous example from Google's self-driving car program: Could your self-driving vehicle recognize someone in a wheelchair chasing a wild turkey down the street as a situation that it should avoid? It will take time for all the computers that run self-driving trucks to store those kinds of possibilities.
Like a lot of critics of educational technology, I hate the frequent analogies that proponents often use to make the major changes they advocate sound perfectly normal. For example, I was just explaining why edtech isn't really like refrigerators or washing machines. Sticking to that same vive la différence vibe, I want to try to explain why — even though teaching isn't really at all like trucking — Madrigal's distinction is still useful.
A few years ago, I spilled an awful lot of pixels over at my blog trying to come to grips with the implications of Massive Open Online Courses (or MOOCs). They were supposed to be the innovation that would not only make most college professors obsolete, but force countless colleges to close as every student would prefer to hear Harvard's best lecture rather than get their course content from the community-college professor in their neighborhood.
Of course, any college professor who cares one whit about teaching understands that education involves a lot more than just conveying information. There's the teaching of particular skills. There's applied learning. There's the unpredictable relationship between two humans whenever they try to to accomplish anything complicated.
In other words, good teaching is just one long series of “edge cases.” You may come into class with the same lecture notes every semester, but unless you spend all your time staring up at the ceiling, how your students interpret the material you're teaching is going to affect the way you choose to teach it. They don't even have to stop you and ask questions while you're talking. So long as you and they are in the same room — with you conveying information in real time — you will see how your material is going over and can adjust your presentation accordingly.
Even if you really could deliver the same exact lecture every time, you will never get the same result twice because the learning process is never entirely predictable. If we automated learning, information would still travel from the brain of the professor to the brain of the student, but we’d never know exactly how well students understood it. You might as well just hit “play” on a tape of someone else's lecture, then leave the room to do something else.
Education shouldn't work like that but shipping goods actually does. Nobody cares exactly how your box of diapers got from its point of production to the place where you bought it. Those diapers will still do their appointed task just the same. But, some will argue: Isn’t the same true for information? Whether you read it in a book or hear it in a MOOC, students still know something that they didn't know otherwise, right?
Only if you have an incredibly narrow definition of education. A definition that narrow should make anyone who's ever been inspired to learn anything truly interesting from any teacher of even limited skill weep like a baby.
If college professors want every day to be the educational equivalent of a milk run, we can teach with the same rote presentation techniques that MOOC purveyors imagine we do. But unlike truck drivers, we can actually use technology to determine our own future. Teachers and professors can employ technology themselves to automate the parts of the process that ought to be automated, and keep the most interesting, unpredictable parts intact. That way, the edge cases that inspired us all to become teachers in the first place will still be around to keep our jobs interesting no matter what technology holds in store.