Comment: If you read about AI, you’ll find a belief that’s widespread: when people don’t want to use AI, it’s because they’re afraid or they just don’t understand it. They’re afraid of losing their job or just afraid of change generally. Or they don’t understand the technology so they can’t see the benefits. They’re stuck in the past.
These contentions frame not using AI as irrational. Intentionally or not, this shuts down real conversations about why people opt not to use the tech and instead makes that choice into a problem to be solved, something to be overcome.
But in our recent paper, we point out there are entirely rational reasons why people might choose to put in the effort of doing something themselves instead of using AI.
AI’s appeal comes down to one big promise: it can save us effort. Take generative AI such as ChatGPT as an example. It makes writing easier, quickly summarises long documents, and can produce images or music on demand. A lot of the arguments for embracing AI are based on being able to do things more efficiently.
If you think about what we do through the lens of efficiency, then adopting AI just seems like the obvious, sensible choice. Surely if you can get the same result with less work, then cutting effort is the rational choice?
This basic idea has driven waves of automation throughout history. Following its logic, using AI starts to look like the only rational option and choosing not to use it starts to look strange, like something that needs to be explained. If you read a lot about AI adoption, you’ll find this attitude is common.
But we think the assumption forgets one simple fact: effort has value, in and of itself. Sometimes effort gets us things that we can’t get any other way. In our paper, we describe three types of value that effort offers—value that can’t be replaced by increased efficiency.
Effort is paradoxical. On one hand, we seek out some activities, such as tricky puzzles or endurance sports, because of the effort involved. On the other hand, we’re probably happy to forgo the effort of doing the dishes. The difference is that some kinds of effort deliver uniquely human value, and we argue that when people choose effort instead of using AI, they’re specifically preserving that value. What kind of value do we mean?
First, effort helps us develop ourselves. If we want to fulfill our potential, and live a full and flourishing life, we need to put effort into mastering things that matter to us and make us better people.
Doing the dishes doesn’t help much with fulfilling our potential so it’s probably the kind of effort we could automate and not miss. But making art, mastering an instrument, or becoming an expert at something allows us to flourish. Even if AI can make a picture very efficiently, we don’t benefit from it if we want to be a painter. We need to put in the effort to develop ourselves.
Second, effort communicates how much we care. Talk is cheap, as they say, and some deeply held feelings are hard to put into words. But putting in effort demonstrates how much something matters.
Think of caring for a spouse during a long illness: it communicates a depth of love that is hard to articulate. And when we outsource effort, for example using AI to write a love letter or note of condolence, even if the words are eloquent, the recipient might reasonably assume that some effort, and some meaning, is missing.
Lastly, whether we like it or not, we live in a society with deep-seated assumptions about the importance of work. We are taught that working hard makes us good people, and that only bad people are lazy or shirk. As such, hard work has moral value. So, when people resist using AI, it might not be fear of change or misunderstanding technology. They might simply feel that avoiding work goes against the grain of entrenched assumptions about right and wrong.
But what about efficiency; isn’t it better to be efficient? It can be, but efficiency is not good in and of itself; we can efficiently achieve awful outcomes. Instead, efficient means are only desirable if the outcome they serve is good. When people resist using AI, rather than fear or ignorance, they may be expressing what they value.
Efficiency is just one thing to value among many. We shouldn’t be willing to trade away other things we value—fulfilling our potential, communicating care, or seeing ourselves as good, hardworking members of society—in exchange for efficiency, at least not without stopping to consider the trade-offs.
This article was originally published on Newsroom.
Rebecca Downes is a senior lecturer in the School of Management and Rebecca Mines is a lecturer in the School of Information Management at Te Herenga Waka—Victoria University of Wellington.