orgtheory.net

why the facebook/okcupid conversation needs the history of technology

I’m in the Poconos this week with old college friends and only intermittently paying attention to the larger world. And I’m hesitant to opine about the latest in the world of online experimentation (see here, here, or here) because honestly, it’s not my issue. I don’t study social media. I don’t have deep answers to questions about the ethics of algorithms, or how we should live with, limit, or reshape digital practices. And plenty of virtual ink has already been spilled by people more knowledgeable about the details of these particular cases.

But I do want to make the case that it’s important to have this conversation at this particular moment. Here is why:

If there’s one thing the history of technology teaches us, it’s that technology is path-dependent, and as a particular technology becomes dominant, the social and material world develop along with it in ways that have a lasting impact.

The QWERTY story, in which an inefficient keyboard layout was created to slow down the users of jam-prone typewriters but long outlasted those machines, may be apocryphal. Perhaps a better example is streetcars.

Historian Kenneth Jackson, in the classic Crabgrass Frontier, showed how U.S. cities were first reshaped by streetcars. Streetcars made it possible to commute some distance from home to work, and helped give dense, well-bounded cities a hub-and-spokes shape, with the spokes made up of rail lines. This was made possible by new technology.

Early in the 20th century, another new technology became widely available: the automobile. The car made suburbanization, in the American sense involving sprawl and highways and a turn away from center cities, possible.

But the car alone was not enough to suburbanize the United States. Jackson’s real contribution was to show how technological developments intersected with 1) cultural responses to the crowded, dirty realities of urban life, and 2) government policies that encouraged both white homeownership and white flight, to create the diffuse, car-dependent American suburbs we know and love. The two evolve together: technological possibilities and social decisions about how to use the technologies. As they lock in, both become harder to change — until the next disruptive technology (((ducking))) comes along.

So what does all this have to do with OKCupid?

The lesson here is that technologies and their uses can evolve in multiple ways. European cities developed very differently from American cities, even though both had access to the same transportation technologies. But there are particular moments, periods of transition, when we start to lock in a set of institutions — normative, legal, organizational — around a developing new technology.

We’re never going to be able to predict all the effects that a particular social decision will have on how we use some technology. Government support of racist red-lining practices is one reason for the white flight that encouraged suburbanization. But even if the 1930s U.S. mortgage policy hadn’t been racist, other aspects of it — for example, making the globally uncommon fixed-rate mortgage the U.S. norm — still would have promoted decentralization and encouraged the car-based suburbs. Some of that was probably unforeseeable.*

But some of it wasn’t. And I can’t help but think that more loud and timely conversation about the decisions and nondecisions the U.S. was making in the early decades of the 20th century might have led the country down a less car-dependent path. Once the decisions are made, though, they become very difficult to change.

Right now, it is 1910. We have the technology to know more about individuals than it has ever been possible to know, and maybe to change their behavior. We don’t know how we’re going to govern that technology. We don’t really know what its effects will be. But this is the time to talk about the possibilities, loudly and repeatedly if necessary. Maybe the effects on online experimentation will turn out to be to be harmless. Maybe just trusting that Facebook and OKCupid aren’t setting us on the wrong path will work out. But I’d hate to think that we unintentionally create a new set of freedom-restricting, inequality-reproducing institutions that look pretty lousy in a few decades just because we didn’t talk enough about what might — or might not — be at stake.

That’s how websites work” isn’t a good enough answer. If 62% of Facebook users don’t even know there’s an algorithm, there’s plenty of room for conversation.

* There is a story that GM drove the streetcars out of business by buying up streetcar companies and then dismantling the streetcars. There are a number of accounts purporting to debunk this story. This version, which splits the difference (GM tried, but it wasn’t a conspiracy, and it was only one of several causes) seems knowledgeable, but I’d love a pointer to an authoritative source on GM’s role.

 

Written by epopp

July 29, 2014 at 7:32 pm

One Response

Subscribe to comments with RSS.

  1. […] to note the different reactions here. Elizabeth Berman treats Facebook and OK Cupid the same way in an excellent post about path dependence. The biggest objections to OK Cupid use their experiment as a way to go back to objections about […]

    Like


Comments are closed.

%d bloggers like this: