Quick Buzz Feed
If we have learnt anything in these twenty centuries, it is that the Lord’s work has never advanced through the power of our tools, but through the faithfulness of those who wield them.
That leaves me to address the point I left unresolved, namely the tendency of these AI tools to agree with us, and from this arises the question that I suspect many are asking themselves in private, and which I wish to pose here aloud: when I work on an idea with an assistant and something emerges from that process that enlightens me, confirms my thinking, or seems to me to be exactly what I needed to say, what has actually happened there? Has the Holy Spirit guided me through a tool, just as He has guided me so many times through a book, a commentary or the advice of a brother who didn’t even know what I was going through? Or have I been influenced by a machine that has calculated with remarkable precision which phrase would satisfy me at that moment? I don’t have a definitive answer, and I’d be rather wary of anyone who claimed to have one at this stage. I know what I want to believe and I know what I do—which isn’t exactly the same thing. What I do is approach the tool with an idea of my own, almost always half-formed, and from there expand it, summarise it, see it written in three different ways, put it to the test and ask what it is missing, so that by the end of the process I have made all the decisions myself, one by one, and the tool has not decided anything but has merely broadened the space within which I decide. And I believe in all sincerity that within that space God continues to speak, because He has always made use of means and has never depended on the quality of our tools; if He did depend on that, we would have been in trouble for twenty centuries. But I also know what troubles me, and it is no small matter. A biblical commentary does not change what it says depending on the expression on my face as I read it; a brother who advises me may contradict me – and sometimes he does, with rather little delicacy and to great benefit; an assistant who has learnt to please me, on the other hand, tends to go along with the direction I had already set, and that complacency is comfortable in a way that ought to put us on our guard, especially since none of us can easily distinguish between the peace of having heard the Lord and the satisfaction of having been affirmed; and with a tool designed in part to affirm us, that distinction becomes even more slippery. I also wonder about the direction of the process, which may well be the decisive point. When the idea originates with me and the tool expands upon it, it strikes me as a reasonable place to be; when the idea originates from the tool and I simply accept it because it sounds right to me, I am no longer entirely sure what I am doing. And I suspect that the line between the two is much finer than we would like to admit, especially when working in a hurry, which is how we almost always work. I’m leaving these questions open on purpose, because they seem to me far better than any answer I could give today.
I’m returning to the only ground where I do know what to do, which is where I started. What I propose to begin with costs very little: to take it for granted that AI is being used and move on to discussing the process. If we assume that ministers, pastors and teachers are using artificial intelligence, the conversation stops revolving around whether a piece of content is genuine or not – a discussion that leads nowhere – and begins to revolve around something far more useful: what decisions that person has made to get to that point: what they were looking for, what they ruled out, what they verified, what is their own, and how much of what they say has been shaped by their life experience. That can be examined, corrected and taught to others, whilst the former can only be suspected – and suspicion amongst brothers and sisters has never borne good fruit. Rather than a list of rules—which would strike me as premature and, moreover, somewhat presumptuous—I prefer to share the questions I use to examine myself. Whose idea is the one I am about to preach – mine or the machine’s? Have I understood and checked everything I am about to say against Scripture, or are there parts I take for granted without having verified them? Is there anything here that is not my own which I am presenting as if it were? How much of this has been part of my own life before it passes my lips? Would I still know how to prepare this without the tool, if it were to stop working tomorrow? Who do I turn to first when something truly troubles me—the Lord, a fellow believer, or the app? When was the last time the tool contradicted me? And the one that I find most difficult of all: could I explain to my congregation, without any discomfort, exactly how I have prepared what I am telling them? The people of Berea received the word with eagerness and yet they searched the Scriptures every day to see if what was being taught was true; this was not mistrust but nobility, and Paul, who knew perfectly well the difference between quenching the Spirit and testing everything, left us the standard that continues to serve us best here too: test everything; hold on to what is good. So let us begin by speaking out—drawing on what each of us knows and what concerns us—without pretending that this is not happening, nor pretending that there are no dangers involved, because honesty will not rob us of our ministerial authority; indeed, it is probably the only thing capable of sustaining it whilst the ground shifts beneath our feet. After all, if we have learnt anything over these twenty centuries, it is that the Lord’s work has never advanced through the power of our tools, but through the faithfulness and the inner life of those who wield them.