Last year I proofread a script for a client’s video ad that compared the experience of using the client’s web design tool to using AI. The ad’s less than a minute long, but sums up an incredibly important principle of how we use and don’t use AI:
Note: Readymag didn’t pay me to share or write this mini-essay, but the video did inspire it.
Besides the actor’s amazing ability to express their impatience using just their hands, and the AI’s overly smug emojis, I like this ad because it focuses on a specific set of circumstances:
- A creative person who knows what they’re doing (“i’m an illustrator with 7 years of experience”)
- They’re panicking about a tight deadline (“the deadline is in 3 days!!”)
- They know what they want to create (“ugh…can you recreate this reference with brighter colors?”)
After attempting to work with the AI interface and give it clear instructions, the creative person gets frustrated and gives up (“this is not working”), telling the AI that they’re looking for something that feels like it came from them and only them (“i need something that reflects my style”). As a result, the ad ends with our irritated creative putting AI away and deciding to make the website themselves.
As a counterexample to this, one of the first AI controversies I saw in early 2023 involved a college-student gymnast and social media influencer who ran a paid post touting AI for writing papers (the original post has since been removed). Notably, the ad’s protagonist is facing a paper deadline “at midnight,” and the ad ends with her successfully producing a finished college essay using AI. Like the ad above, both users turn to AI because they’re facing an immediate deadline—and, presumably, all of the nail-biting anxiety and fears that go along with it.
On a bigger level, the use of urgent deadlines in these ads seems to tap into a phenomenon of people turning to AI not only because they’re stressed over finishing one particular task, but because they’re rushed for time in general and have too many things to do. The presumption is that AI will help them accomplish their tasks quickly and easily as a labor-saving device.
On a broader scale, though, both ads suggest the idea of fear: fear that we as humans won’t be able to accomplish our goals. These can be either goals we’ve set for ourselves, like the illustrator applying for her residency, or goals that the outside world has thrust on us, like the college student writing her paper. The second ad, like countless other ads across decades, seeks to tap into basic human fears that something in our lives will go wrong, and that simply shelling out some money for the product in question (or in this case, using AI) will make the problem go away.
Beyond this, though, the protagonist of the college-paper-writing ad presumably has no interest in the content or finished product of her essay. Unlike the illustrator in the first ad, her goal isn’t to produce something interesting, something that other people will find valuable, or something that reflects her individual talent. No: her goal is to meet her deadline.
While the first ad shows us the generic gallery-type websites the AI generates for the illustrator, the second ad never really shows us the gymnast’s AI paper, so we don’t know the quality for sure. However, common experience with AI-generated slop, suggests that the paper will be generic and uninteresting depending on the amount of time she takes to review and edit it.
This brings me to my bigger point: the protagonist of the first ad realizes that the work AI is generating for her is generic and uninteresting, so she decides to make the website herself. The protagonist of the second ad doesn’t care that the work AI is generating for her is generic and uninteresting, so she turns it in and goes to bed satisfied.
Do Your Own Work When It Matters (and Maybe Use AI When It Doesn’t?)
The comparison above seems to be pointing toward what I believe to be the future of AI use—that humans will continue to create their own work in situations that are meaningful, either to themselves or to others, and will use AI for work where they just don’t give a fuck.
Let’s face it: we all do plenty of work where we don’t really care about the outcome or learn anything in the process: we just need to achieve a result. Most people don’t gain any self-benefit from washing dishes by hand, and no one else seems to benefit from you washing dishes by hand either, so why not just use a dishwasher and accomplish the same result (or better) with less effort?
Setting aside arguments about the cost of buying and owning a dishwasher, plus differences in the costs of dish soap and the mental and physical benefits of doing a physical chore like washing dishes, my dishwashing example above is one of many examples in our lives where we look to machines to accomplish tasks we don’t want to do, and where the result is basically the same either way. As long as the dishes get clean, we’re happy.
Consider the two ad examples above one more time. In the second one, the college student is happy with her AI-generated result. This suggests that she doesn’t care about writing an interesting paper, engaging with her teacher or classmates, or developing her paperwriting skills. Her goal is a set result: to finish the paper. And as long as the paper is finished, she’s happy.
Now consider the first example again. Our protagonist is an illustrator with seven years experience who knows what she wants and tries to guide the AI toward that goal using specific feedback, and she grows visibly frustrated when the AI can’t give her what the quality end result she’s imagining. She cares about making a great-looking website that conforms to both her personal aesthetic and reflects her as a professional who doesn’t want to share mediocre work with the world. For her, it’s about more than just achieving a result—it’s about achieving a certain kind of result that feels like the best it can be and reflects her own style, both to the people around her, and to herself.
In other words, the protagonist in the first ads cares about her work, and the protagonist in the second ad doesn’t.
Maybe Our Use of AI Says More About Us Than We Realize
I think a lot about the effort I put into things, both the things I care greatly about, and the things I just kind of do in what my eighth-grade math teacher used to call a do-the-job way. I put vastly more effort into work I genuinely care about because that work represents who I am as a person—and ideally, I’d like that to be all the work I do. But for work I care less about or really don’t want to do, I tend to put forth far less effort.
Maybe instead of getting excited about AI and finding more and more ways to use it, we as a society should first work on separating tasks that matter from tasks that don’t matter. Maybe then we can start openly encouraging (or even streamlining) AI for the tasks that don’t matter or don’t possess intrinsic value, and start openly condemning the use of AI for tasks that do matter, since using AI for these tasks just produces more AI slop that no one wants to see, replaces human-authorship, and makes society in general a less interesting place.
In general, I’d like to see everyone focusing more on the things we actually give a fuck about and want to remain human-authored while relegating everything else to the status of dishwashing: using AI and machines to do those things so we can do the meaningful work ourselves instead.
Like this post? Sign up for my email list to get future posts sent to your inbox.
