Why do we pay attention when someone shouts?
It’s not only because the sound is loud – not only because it blocks other sounds and hurts our ears.
It’s not only because there’s a social norm that shouting indicates urgency.
It’s also because shouting takes effort. A person can’t shout for very long before getting tired and hoarse. So someone who shouts is making an investment, and we sense that.
If our physiology allowed us to shout with the same ease as we speak, then shouting would lose significance as a gesture. We might still notice it, in the way we notice an emoji, but we wouldn’t take it as seriously.
This example reveals that when we communicate, we do more than exchange information with each other. We model each other. We imagine what it might take for the other person to make the gestures they’re making. Our empathetic model of the other person affects how we experience their words and actions.
It’s always been possible to game the system of empathetic modelling, so to speak. For example, some people just naturally have loud voices. They’re not putting in extra effort to speak so loud, but we respond to them as if they were more passionate than the quieter ones. Those who speak softly are at a disadvantage when it comes to garnering attention, even though they might be the ones trying harder.
Until recently, writing was hard. The only way someone could produce a large body of text that exhibited good prose style and good logic was through time and effort. Some people were much more fluent at writing than others, but we could assume that any person who wrote a tight essay or a well-organized book must have cared about what they were writing, or at least, they must have had a motivation significant enough to compel them to invest the time and energy required.
Now it turns out that computers can write for us, and they can write better than many of us. So how will this affect our empathetic modelling of the written word? How will it affect our ability to be moved by what we read?
Surely we will still find meaning in text, but there will be a cloud of uncertainty about the purported author’s investment in the act of creating that text. Did they spend hours laboring over a from-scratch essay or did they just lightly revise the near-instantaneous product of a machine?
Already some writers are saying “AI is my partner. It helps me express myself.” If the writing is good, informative, entertaining, why should it matter how it was produced? It should matter because communication is more than an exchange of information. Communication is an experience, and that experience can’t be detached from the context – from our understanding of the properties of the medium, including its difficulty.
If we want a glimpse of how this is going to go, we can look to photography. The internet is awash in photographs and we can still enjoy them and learn from them and sometimes be moved by them. But we can never be sure how they’re made. And that devalues them. A digital photographer might have labored for months to get a certain shot – their dedicated practice finally combining with luck to achieve a miraculous result – or they might have taken a lackluster image and enhanced it by clicking a few buttons in photo-editing software to achieve that marvelous color and detail. Or maybe they just typed a description of what they wanted and had the software generate the photo for them?
If it’s a good photo, who cares? We care because we want to be moved, and our ability to be moved depends on our ability to connect with the artist by imagining their process of creation. Empathetic modelling.
The prerequisite to being moved is having our attention captured. And our attention is captured when we see evidence of effort. Proof of work. Demonstration of commitment.
With technology, we’re making things easier and easier so that no creative product can any more be seen as proof of work. Awe at an artist’s labor is replaced with a question mark. How did they do it and how much help did they have from machines?
Is there something morally wrong with getting help from a machine? That is not the question at hand. The question is what happens to aesthetic experience when the process of creation becomes increasingly machine-driven? The question is what happens to our response to a communication act when empathetic modelling is infused with doubt, when we can no longer discern what an author wrote and what they merely took? What happens to aesthetic experience when machine-brokered magic is inserted into our image of the process of human creativity?
Our obsession with authenticity should tell us something about this. Why would we gaze for hours at an old masterwork hanging in a museum but not at a forgery that looks identical to it? That’s because when we see the authentic work, knowing that it is such, we imagine the great artist laboring to make it, but when we see the forgery, we image a thief working to deceive us. Same image, different empathetic modelling, different experiential outcome.
If we look back to the time before machines could write, a time when photographs could only be made with light hitting physical film, it wasn’t a time of unbridled bliss. Glorious words could still be ignored or misunderstood, and photographs themselves may have seemed too easy to make in comparison to paintings, too easy to be worth a viewer’s deepest respect. But anyone who tried to take a photograph would have come to know the difficulty of it and been able to appreciate the accomplishments attained in the best photographs.
Imagine a world where we’re all shouting, all the time, and we can’t be moved by any of it. That is the world we would get if, through technology, we made it effortless to shout. ■