A Question for You
Earlier this week, I was spitballing with a friend about possible ways to keep dangerous people out of power, regardless of party:
“What if we had, say, a Dehumanization Index? Where the volume of dehumanizing public statements a person has made follows them around—as a known résumé feature, much like a previous job or educational credential?”
My thought process was this:
Dehumanizing statements are fundamentally different from common gaffes. Almost every public figure will say something stupid at some point, but those who have a pattern of saying dehumanizing things are revealing something important: they have the capacity to erase or diminish the humanity of other people, whether out of genuine belief in their own superiority, political expediency, or both. No matter the motivation, dehumanizing language is dangerous in a way that other language is not. We should allow it—free speech and all—but we should judge it.
The capacity to diminish the humanity of other people may be more widespread than I’d like to think, but if there’s one place it shouldn’t be common—in fact where it should be explicitly disqualifying—it is political leadership. Especially as AI proceeds to undercut the economic value of human labor, the intrinsic value of every human being needs not just to be seen; it needs to be championed.
“Nah,” said my friend, responding to my index idea. “Two problems: 1) It’s too subjective. One person’s dehumanizing statement is another person’s “truth telling”; it would be impossible to establish a broadly-accepted, non-partisan index. 2) Even if we did manage to arrive at a single index, it wouldn’t move the needle for voters. Trump voters knew what he’d said—and they voted for him anyway.”
“Dude,” I replied, head in hands. “That is the least ‘Yes…and’ response ever. We’re trying to repair the world here. Let’s workshop this.”
I went on: “When people see a sign in a restaurant window with a failing grade for food safety, they don’t ask themselves, ‘Am I a Republican, or am I a Democrat?’ They don’t want food poisoning, so they pick a less risky place with a similar offering. Imagine that I’m focused on tax policy; the index might help me choose a candidate whose tax ideas I like—without also having to subscribe to a dark and dangerous worldview.”
I kept going: “I’m not saying everyone would be influenced by the Dehumanization Index—the world will always have some people who prefer to make decisions in an information vacuum—but I think enough people might be influenced to make a difference. And while it used to be marginalized groups who fretted most about dehumanization, AGI has the potential to marginalize us all. I think these days we could get a lot of people caring about our index.”
At this point, my friend had his head in his hands. But I forged on: “As for the subjectivity issue,” I said, “let’s use the technology! There is a tremendous amount of scholarship, from people all along the political spectrum, regarding what constitutes dehumanizing language. So, we create a non-partisan rubric, and we use twenty different tools to land on an aggregate score that is as devoid of bias as possible. C’mon, you’re the math guy. WORK THE PROBLEM.”
We had to go our separate ways at that point, so now I turn to you:
If we had a way of scoring our potential leaders not on a fuzzy parameter like “character” but on something quantifiable—how often has this person publicly engaged in the specific act of dehumanizing others—could it help us sort for the kinds of leaders this moment demands? Could it help us sort for…basic human decency?
Notes:
Hat tip to my uncle, Peter Elliott. Before I had this conversation with my friend, an email exchange with Peter got me thinking more about how exactly one might assess someone’s character from afar. (Peter writes beautifully about art history, by the way. His six books, including his newest, Looted!, are listed here.)
