Artifacts Have Politics. Now What?

I’ve read Langdon Winner‘s essay “Do Artifacts Have Politics?” a dozen or more times. I first read it in grad school in the 1990s, and now I assign it in… well, almost ever class I teach.  Winner shows that some artifacts have deliberate politics. The highway overpasses around New York City were deliberately designed to keep poor people (especially non-whites) away from the beaches.  We know this because their designer, Robert Moses, said so.  Other artifacts (like nuclear versus solar power)  are not necessarily intentionally political, but lend themselves to certain kinds of power arrangements.

I once attended a lecture Winner gave, and during the question period asked him: “OK, I’m an engineer and I accept everything you say. What would you like my peers and I to do differently?” He didn’t really have an answer. I guess it’s kind of a hard question.  And I’m still pondering it myself.  I suppose “be mindful” is one straightforward answer, but the details matter–and the details in practice aren’t obvious.

When I teach the paper, I often use face recognition technology as a discussion topic. If you could invent perfect face recognition, would you?  If for example you could set up a camera at every convenience store and gas station in the nation that would reliably identify bomber Eric Rudolph while he was on the loose, would you? Are the implications different if, as is inevitable, the technology has an error rate? This leads to a discussion of the checks and balances we have in US law and whether we really trust the government to honor them in practice. If we err on deciding that we will trust the government and work within the system to make sure the limits are respected, that leads to a scarier question: What about use of this technology by totalitarian regimes in other nations?  If you invented it, wouldn’t they eventually get access too?  Is the inventor responsible for all of a technology’s eventual uses?  Knowing this, would you want to be the inventor or not?  It’s reasonable to say you’ll invent it and try to stay involved in the broader sociopolitical context of its use in practice–maybe that’s really all you can do. But you also need to recognize that you will sometimes lose control of what happens next.  Which raises the question, is there any technology that, given that you will likely eventually lose control of its uses, that you would decline to invent?  It’s easy to come up with an absurd example where the answer is yes (Marvin the Martian’s “destroy earth” button comes to mind). It’s harder to come up with heuristics for when something less extreme might fit that description.

I’ve had this discussion with classes over and over. It’s a great conversation–it gets students thinking.  And the discussion time and time again has followed the same path–until yesterday.  In my “Intro to Human-Centered Computing” class yesterday, master’s  student Vincent Martin commented, “I need face recognition technology. I would love to be able to recognize my friends and family again.”  Vincent is blind.  I’m surprised that obvious application doesn’t come up in conversation every time we discuss this issue.  I’ll make sure it comes up in the future.  If you were at all leaning towards refraining from developing face rccognition technology, I hope this would change your mind.  For every basic technology we develop, how many hidden surprise–uses for both good and ill that we can’t anticipate–are there?

OK, artifacts have politics. What does this mean for us as designers and engineers? Beyond high-level platitudes like “be mindful,” what should we do differently? I’m still wondering.