more on the Liahona, efficiency, and technique

Yesterday afternoon, I was explaining (poorly) to some friends that I had been thinking about what the story of the Liahona in the Book of Mormon has to teach readers of that volume of scripture about (generative) AI. So, that connection was naturally on my mind when I was reading more of Jacques Ellul’s Presence in the Modern World over breakfast.

I continue to be pleasantly surprised by how relevant Ellul’s writing feels for today. Presence in the Modern World was first written in 1948, and even if you consider that the translation I’m reading is based on a 1988 second edition, that’s still enough time to earn the description “prescient.” (I’ve been reading Ellul in a mix of translated English and original French, depending on what’s more practical for the book in question.) Here’s a passage that particularly stood out to me this morning:

In reality, what justifies the means today is whatever succeeds. Whatever is effective, whatever possesses in itself an “efficiency,” is justified. By applying means, a result is produced. This result is judged by these simplistic criteria of “more”: larger, faster, more precise, and so on. Simply by applying this criterion, the means is declared good. What succeeds is good, what fails is bad. Now, technique teaches us how to infallibly discern the means, the one means that carries within it the success that is most spectacular. Technique always succeeds. The most-perfected technical means attains necessarily all technical objectives (which are not ends—confusion must be carefully avoided).

I’ve written a few times (and thought countless times more) about how relevant the fetishization of efficiency is for the 2020s, but one more Ellul excerpt to cite on that point is, of course, not unwelcome. One thing that really stands out from the passage, though, is in the latter half: that success is what matters, but that success is not necessarily measured in ends but in means.

So, yes, the CEO of Duolingo is now claiming that AI is better than teachers (a seeming appeal to ends), but the CEO’s claims may actually be more rooted in means:

The Duolingo model of teaching via quizzes and drills isn’t suitable for all subjects, he said, noting that history might be a subject better taught with “well-produced videos”—something AI can’t currently do well. But he said he believes the problem of scale tips the balance on the side of AI.

If “it’s one teacher and like 30 students, each teacher cannot give individualized attention to each student,” he said. “But the computer can. And really, the computer can actually … have very precise knowledge about what you, what this one student is good at and bad at.”

Stories like the one about the Chicago Sun-Times summer reading list also demonstrate how much means rather than ends are the real success of AI. Yes, it’s super embarrassing that an insert included a bunch of fabricated information, but since the insert came from King Features syndicate, and since the business model of a newspaper syndicate is to provide a bunch of content to a bunch of newspapers (which means that more efficient means leads to a better bottom line), I am skeptical that companies like this will back off of exploring generative AI. Yeah, these ends were straight up wrong, but the means were still damn good in terms of profit, and as long as there aren’t enough embarrassing ends to cut into that profit, I expect the means to win out.

Anyway, back to the Liahona. Ellul bemoans that results are “judged by these simplistic criteria of ‘more’: larger, faster, more precise, and so on.” It’s kind of remarkable, though, that by these standards, the Liahona fails every time. As I suggested in my last post, the weird thing about the Liahona is that it is introduced into a narrative that has already established that those using it have previously spoken with God (through visions and dreams) and received visits from angels. Does this weird little machine provide a larger outcome than a visit from an angel? Is it faster than a vision from God? Is it more precise than revelation coming through a dream?

I’m skeptical that anyone could compellingly argue for an affirmative answer to each of those questions, and while it’s fun to ask the Star Trek V question “what does God need with a Liahona,” I’m more interested in using this obvious lack of efficiency to develop a theology of technology for Restoration traditions. In Anarchie et christianisme (Anarchy and Christianity), Ellul argues:

Si le Dieu biblique est le Tout-Puissant, il est en même temps celui qui pratiquement ne se sert jamais de sa Toute-Puissance dans sa relation avec l’homme[.]

That is (with a couple of editorial translation choices),

If the God of the Bible is all powerful, this God is also one who practically never uses that power in their relationship with humanity.

See also this observation in La parole humiliée (Humiliation of the Word):

Car le pouvoir de Dieu est un pouvoir retenu, Dieu n’occupe pas toute la place.

Or:

For the power of God is a retained power, God does not take up all available space.

I like this take, even if have some quibbles with the details (there are times where the biblical, or Book of Mormon, God does exercise power to horrific ends, though maybe that just makes it more important to focus on a God of “non-power”). So, along these lines, and like I’ve written previously, maybe what the seeming uselessness of the Liahona can teach readers of the Book of Mormon is that God is also largely disinterested in efficiency, in the fetishization of means. (Ellul connects power and his broader concept of technique in Théologie et technique, though I’m not going to go hunting for a quote right now).

It seems to me that a God who uses a Liahona to communicate with humanity would be, at most, indifferent toward AI and more likely skeptical. After two blog posts on the subject, I’m even more interested in fleshing this argument out more, and I hope I find an opportunity to write it up in the future.

• Liahona
• technique
• technology
• Jacques Ellul
• Book of Mormon
• AI
• generative AI
• Presence in the Modern World
• Anarchie et christianisme
• La parole humiliée
• non-power
• Théologie et technique