It’s no secret that I’m not a fan of generative AI. I can feel my colleagues’ eyes rolling before the words “but the copyright infringement and environmental impact” slip from my lips yet again. (Well, not all my colleagues. Some are staunchly anti-GenAI, as well.) I have no interest in using it to “make my job easier” or to write things for me – or to have it hallucinate incorrect answers to research questions that I would have otherwise rabbit holed down. (I mean, seriously, stop trying to take away my joy.)
Then I read an opinion piece that added to my list of reasons of why I will continue not to use GenAI: busywork.
According to this piece, more and more people (at a seriously alarming rate) are turning to GenAI to ‘help’ them with things. According to the piece, people are feeling more independent. But in reality, AI is creating a massive transfer of labor.
You have my attention.
Consider this analogy provided by the author Dr. Carl Benedikt Frey – the washing machine. In the 1800s, laundering was an occupation (mostly in cities and for people who could afford to pay someone to do their laundry). This was a lot of work: “hauling water, chopping fuel, boiling linens in lye, scrubbing each garment by hand against a washboard, wringing, drying, starching and ironing with heavy flat irons heated on a stove.” A lot of work – and it was all paid labor.
Enter the washing machine. It changed laundering – but it did not disappear the labor. It instead shifted it, unpaid, onto the household – “The laundress lost her job. The housewife gained a chore.”
This is a pattern we can see repeating over time. Consider self-checks at stores. Consider the last time you consulted a travel agent (ask youngsters what a travel agent even is…). You don’t even need to visit a bank to deposit a check anymore. Each of these shifts the labor (again, unpaid) onto you. Not a big deal, no.
Until you realize that people are using platforms like Ch@tGPT to answer medical questions or even do their taxes.*
There is nuance in both those examples that an LLM just cannot replicate. But also – it’s shifting that labor, unpaid and (more important/scary) unskilled, onto us. We are doing more with less. While we may not notice the additional labor when we scan and pack our own groceries, these smaller tasks add up. As the opinion piece points out, there’s a name for that: “opportunity cost neglect — the well-documented tendency to overlook the value of what we give up when the cost is time rather than money.” And maybe the time is worth painting that fence rather than paying someone to do it. But asking Ch@tGPT how to paint that fence? Well, we need a new word that talks about “opportunity cost neglect” in addition to “extra resources wasted when there are already other options available to us.”
Making the choice to add this busywork (i.e. unpaid labor) onto ourselves, we do change the way the world works. No one is measuring the time you spend scanning your groceries or doing your taxes at home or painting that fence. As more of us do these things, the labor stats disappear, along with the jobs. Corporate profits go up – and we’re left feeling overburdened. From the opinion piece: “The laundress disappeared from the statistics long before she disappeared from memory. Many more trades and professions are on the verge of the same shift. The A.I. revolution may not have taken your job yet. But it has already put you to work.”
(*Some people will say that platforms like Ch@tGPT are democratizing because not everyone can afford to pay a professional. BUT. We don’t need to ask Ch@tGPT what we can make with the ingredients in our fridge – there’s already an app for that. We don’t need to ask it how to fix our fridge’s ice machine – there are already YouTube videos by experts that can walk us through that. Ch@tGPT is not solving new problems. It’s poorly attempting to solve problems we already had solutions for – and doing so using more energy/resources and built on the theft of other people’s work.)
Ami Maxine 