Why does AltSchool, as described in this article, as well as similar kinds of tech-industry attempts to “disrupt” education, bug me so much? I’d like to be more welcoming and enthusiastic. It’s just that I don’t think there’s enough experimentation and innovation in these projects, rather than there being too much.
The problem here is that the tech folks continue to think (or at least pretend) that algorithmic culture is delivering more than it actually is in the domains where it has already succeeded. What tech has really delivered is mostly just the removal of transactional middlemen (and of course added new transactional middlemen–the network Uber has established in a really frictionless world wouldn’t need Uber, and we’d all just be monetizing our daily drives on an individual-to-individual basis).
Algorithmic culture isn’t semantically aware yet. When it seems to be, it’s largely a kind of sleight-of-hand, a leveraging and relabelling of human attention or it is computational brute-forcing of delicate tasks that our existing bodies and minds handle easily, the equivalent of trying to use a sledge hammer to open a door. Sure, it works, but you’re not using that door again, and by the way, try the doorknob with your hand next time.
I’m absolutely in agreement that children should be educated for the world they live in, developing skills that matter. I’m also in agreement that it’s a good time for radical experiments in education, many of them leveraging information technology at new ways. But the problem is that the tech industry has sold itself on the idea that what it does primarily is remove the need for labor costs in labor-intensive industries, which just isn’t true for the most part. It’s only true when it’s true for jobs that were (or still are) rote and routinized, or that were deliberate inefficiencies created by middlemen. Or that tech will solve problems that are intrinsic to the capabilities of a human being in a human body.
So at the point in the article where I see the promise that tech will overcome the divided attention of a humane teacher, I both laugh and shudder. I laugh because it’s the usual tech-sector attempt to pretend that inadequate existing tech will become superbly useful tech in the near-term future simply because we’ve identified a need for it to be (Steve Jobs reality distortion field engaged) and I shudder because I know what will happen when they keep trying.
The central scenario in the article is this: you build a relatively small class with a relatively well-trained, attentive, human teacher at the center of it. So far so good! But the tech, ah the tech. That’s there so that the teacher never has to experience the complicated decision paths that teachers presently experience even in somewhat small classes. Right now a teacher has to decide sometimes in a day which students will get the lion’s share of the attention, has to rob Peter to pay Paul. We can’t have that in a world where every student should get all the attention all the time! (If nothing else, that expectation is an absolutely crystallized example of how the new tech-industry wealthy hate public goods so very much: they do not believe that they should ever have to defer their own needs or satisfactions to someone else. The notion that sociality itself, in any society, requires deferring to the needs of others and subsuming one own needs, even for a moment, is foreign to them.)
So the article speculates: we’ll have facial recognition software videotaping the groups that the teacher isn’t working with, and the software will know which face to look at and how to compress four hours of experience into a thirty-minute summary to be reviewed later, and it will also know when there are really important individual moments that need to be reviewed at depth.
Here’s what will really happen: there will be four hours of tape made by an essentially dumb webcam and the teacher will be required to watch it all for no additional compensation. One teacher will suddenly not be teaching 9-5 and making do as humans must, being social as we must. That teacher will be asked to review and react to twelve or fourteen or sixteen hours of classroom experience just so the school can pretend that every pupil got exquisitely personal, semantically sensitive attention. The teacher will be sending clips and materials to every parent so that this pretense can be kept up. When the teacher crumbles under the strain, the review will be outsourced, and someone in a silicon sweat shop in Malaysia will be picking out random clips from the classroom feed to send to parents. Who probably won’t suspect, at least for a while, that the clips are effectively random or even nonsensical.
When the teacher isn’t physically present to engage a student, the software that’s supposed to attend to the individual development of every student will have as much individual, humane attention to students as Facebook has to me. That is to say, Facebook’s algorithms know what I do (how often I’m on, what I look at, what I tend to click on, when I respond) and it tries (oh, how it tries!) to give me more of what I seem to do. But if I were trying to learn through Facebook, what I need is not what I do but what I don’t! Facebook can only show me a mirror at best; a teacher has to show a student a door. On Facebook, the only way I could find a door is for other people–my small crowd of people–to show me one.
Which probably another way that AltSchool will pretend to be more than it can be, the same way all algorithmic culture does–to leverage a world full of knowing people in order to create the Oz-like illusion that the tools and software provided by the tech middleman are what is creating the knowledge.
Our children will not be raised by wolves in the forest, but by anonymously posted questions answered on a message board by a mixture of generous savants, bored trolls and speculative pedophiles.