The Ideological Turing Test

I recently published an essay called Your Information Diet that summarizes my frustration, informed/inspired by the 10+ years I spent immersed in the world of competitive debate, about the current state of political discourse in America (and quite possibly elsewhere – I’m not qualified to speak to that one way or the other). It’s about the importance of challenging your views, not simply with any old ideological opponent, but with the most compelling advocates of opposing views.

One source I’ve found useful for offering a credible challenge to my thinking is Tyler Cowen, who is thought-provoking both in his own writing and with his curation of guests for his podcast. I don’t think Cowen and I have wildly opposed worldviews, and in fact much of his value isn’t so much in challenging what I already think as it is in suggesting new things I hadn’t thought about at all. However, I do think that he and his guests tend to be more optimistic about technology and capitalism than I am, and they also know more about those subjects, so they make for a good read/listen.

As it happens, Cowen’s most recent podcast guest, Bryan Caplan, actually during their interview on the subject of understanding opposing viewpoints. Drawing on John Stuart Mill in much the same manner that my essay he does, he suggests the Ideological Turing Test to assess how well one understands the arguments of those with whom one disagrees ideologically. The idea is that, much like the Turing Test for artificial intelligence, a person who truly understands an ideological viewpoint with which he disagrees should be able to answer questions in a way that is indistinguishable from a person who agrees with that ideological viewpoint. In other words, a person who understands, say, Hayekian libertarianism should be able to answer questions in exactly the same way that an actual proponent of that ideology would.

The one thing that this Caplan’s test doesn’t account for explicitly, which is a central point in my essay, is which proponents of an ideology one should aim understand well enough to mimic. Even if I could pass for Ann Coulter – a person whose espoused views on immigration differ sharply from my own – in an Ideological Turing Test, I imagine there are a lot of people who might broadly agree with Coulter in terms of policy outcomes but still not consider themselves ideologically identical to her. Perhaps that’s just a semantic point, though.

Regardless, I think switch-sides debate (any format that requires participants to argue both sides of a topic over the course of a competition) is a pretty good real-world analog to Caplan’s Ideological Turing Test. There has actually been a trend of debaters trying to find various ways around arguing both sides of a topic, and while I’m broadly inclined to find that problematic, I also must admit that I probably wouldn’t pass a Turing Test if I had to provide you with the arguments in favor of those endeavors. So, I try to reserve judgment on that issue unless it arises in a debate that I’m adjudicating, in which case the objective as with any issue is to reach a conclusion based on the arguments presented during the debate rather than based on my own preconceptions.

By the way, shoutout to Nate Meyvis for introducing me to Cowen in the first place.

1 thought on “The Ideological Turing Test”

  1. I definitely thought of you during that part of Caplan’s interview.

    I also think some familiarity with economics is a great way to construct arguments that oppose the ones you’re used to. It provides a lot of tools for considering alternative causes, for thinking about morally charged topics from a less moralized perspective, and for fitting a phenomenon into different frameworks.

    Whether or not that’s plausible to you, I’m sure you know by now that Tyler is great at coming up with a dozen arguments on each side of the issue.

    Just one misc., slightly related thought here, because the thread has come and gone: I agree that Scott Adams’ Trumpian writings are full of terrible arguments and often just annoying, but he really does occasionally capture something about what’s compelling about Trump. He was one of the first people that caused me to think that, beyond party affiliation / racist sentiment / etc., Trump represented or communicated something that resonated with a lot of people. I’d have a hard time pointing you to a specific line of his that does this, but if I read a few hundred words of Adams while pretending that Trump is some foreign politician I don’t really know anything about, I find I come away from it with a slightly better perspective on Trumpism.

Comments are closed.