The price of process

(Photo by Natalie Cardona on Unsplash)

Like Maren, I read David Sax’s The Revenge of Analog last month, and some points in it chimed with some other thoughts I’d been having around AI. The book makes the case around how analogue industries and formats have revived despite their apparent inevitable demise in face of digital alternatives. It is sometimes too keen to reinforce its won hypothesis and ignores counter points (the education chapter had me wincing in places for over-simplification), but overall it marks an interesting reaction to technology. It can be viewed in some respects as an argument against technological determinism, that despite all of these predictions of doom, people tend to behave in unexpected ways and new models come through.

This is not just persistence of the old format, but how that format or practice is changed through the basis of a digital prevalence. For example, the sale of vinyl records has grown steadily over the past 20 years but that doesn’t mean the people who purchase them don’t also have Spotify accounts, and use this to discover new music, or that artists don’t use digital means to connect with audiences and distribute music.

He also makes the point that analogue is often more profitable than digital, whether it be sales or advertising, it is a case of “digital pennies versus print pounds”. And this gets to the feeling I had about AI following on from the inevitable discussions at conferences about its impact. The areas where analogue has proven resistant and popular is where people have come to value the process over the product. Tabletop gaming is a good example – many of the games could be played as well, if not better, online but it is the process of sitting around a table with friends, interacting over the game that is valuable. And while audiophiles make a claim about the sound quality of vinyl, I think it is really the process of purchasing, owning, handling and listening to music in a physical format that is the essence cherished by those of us with unwieldy vinyl collections.

Which brings us on to AI. The use of AI tools is generally fine when the process itself doesn’t matter. Producing monthly reports that no-one will read? Sure, use AI. Making an image for a presentation to a few people? DaVinci your heart out. However, if you are writing a paper, or an essay where the act of writing increases your own understanding then AI is at best a prompt. If you want to create art for a research project that reflects your values and identity then the act of doing so (as we know with our work with Bryan Mathers, who always emphasises that it is the conversation that is important) is in itself valuable. Plus you are demonstrating that it matters to you by employing someone or utilising your own time.

What this perspective reveals is how much of the higher education experience we have made seem like it doesn’t matter. Writing research papers is merely a means to increase an h-index, or get promotion, so you may as well get AI to do it. We have stressed grading to such an extent that writing essays for students is mainly only about getting the grades, so why shouldn’t they use ChatGPT? The focus on performance metrics has created a sector filled with garbage process that we have to undertake because that is how to win but that are essentially meaningless in themselves. They are begging to be AI’d to death.

Ted Gioia has an interesting newsletter post about MacGuffins. The point of a MacGuffin in film is to give the onscreen character some reason to pursue their quest, the Holy Grail being the archetype. But actually he notes, the MacGuffin is in itself irrelevant, it could be anything. When it becomes a physical object it becomes meaningless “The quest was previously about transforming your life. Now it gets turned into a physical object—and a vague one, with all the key details missing.” This could be a description of much of higher ed.

There is something in the value we place upon certain analogue based experiences and higher education’s response to AI I think. It should be a wake-up call to reinstate the value of the process itself and to consider the importance we place upon garbage processes. It’s worth asking “does this process matter?” if the answer is no, then expect it to fall foul of AI. If the answer is yes, then we need to demonstrate and reinforce that, because it is difficult to tell now which of the processes actually add value to the individual undertaking them, and which are just meat for the metric grinder.

This work is licensed under a Creative Commons Attribution 4.0 International