"That's So A.I.!"

I’ve made no secrets about my frustrations with A.I. before. Whether it’s writing several articles on it, or lamenting how it shouldn’t detract from people’s livelihoods, A.I.’s become one of my hot-button topics. However, with that comes the flip side of people attributing something to A.I. that shouldn’t be. This is especially the case with the phrase “that’s so A.I.” among younger adults, such that it’s disparaging art for the wrong reasons. It’s tiresome.

Take the latest season of Stranger Things. I happened to love it, warts and all, but there’s been a sticking point since the Netflix documentary debuted on its production. One of the big complaints was how the Duffers “didn’t have a finished ending when they started filming”, hence they “relied on ChatGPT”. I must ask if the complainers have actually written for television, or even a story in general, because that’s not how it works. Deadlines exist, and since scenes are often shot out-of-order, it’s not uncommon to start with production before the final episode’s written. That also doesn’t mean the show-runners don’t know the ending.

I take offence to this particular claim because I’m a writer. I’ve been working on novel manuscripts for several years, and one of them is currently being worked on with a professional editor. Writing stories is more collaborative than people think, and that involves incorporating feedback. If you want proof, I revised a character’s arc based on feedback as I was in the process of writing. It also made the story better.

By pinning the flaws of Stranger Things on A.I., people are reading too deeply and not thinking clearly. Sometimes, something’s messily-written for no other reason than it was messily-written. There doesn’t need to be a profound explanation. It also detracts from the bigger issue, that being how A.I.’s used in sneakier, more noticeably-awful ways. Simply fool around with ChatGPT to see that.

 

Another example involves The MCU’s Phase 4 and 5 offerings. Many people weren’t so big on them, calling the movies and shows “A.I. created” like that absolved them of critical thought. Ignoring how the only instance I can think of A.I. use involves the title sequence of Secret Invasion, associating the flaws with A.I. ignores how numerous factors could’ve made them less-than-optimal. Like how Marvel was releasing so many shows and movies a year that not enough time was dedicated to each one.

Perhaps the most-egregious example is Wish. I thought it was fine, if uninspired, but that’s an example of A.I. potentially impacting quality. How can I tell? Because the song lyrics have syntax errors. There’s also the issue of tight deadlines here, since Disney releases movies yearly, but A.I. was clearly used to write these songs. Simply put, the movie wanted to copy Lin-Manuel Miranda’s style, but failed. And it used A.I., to noticeable effect. That’s a problem, one that should be called out even if the movie was mediocre in every other way.

This is the problem with the A.I. claim: it excuses human folly. Storytelling isn’t easy, especially in a visual medium. Plenty of hands touch even simple projects, and so much could go wrong at any point. Truthfully, if even a bad movie or show makes it to completion, that’s an accomplishment. Because for all the ones that are finished, there are so many that don’t see the light of day.

Complaining that something you don’t like is “A.I.”, especially when it’s not, does a disservice to the artists. Even talented people make bad art occasionally. It’s a byproduct of being human, and that’s okay! If we’re to improve our artistic capabilities, then we need the potential to misfire, sometimes spectacularly. We also need to learn and grow from that.

I’m not saying A.I. doesn’t occasionally sneak into the experience. However, there are ways of noticing that, like how The Book of Boba Fett used deepfake technology with young Luke Skywalker’s face and voice. That’s genuine A.I., yet it didn’t bother many Star Wars fans because “Luke Skywalker was back”. Ideally, we’d call that out more. But that’s asking too much from a fandom that sends death threats to creatives for being adventurous…

Perhaps the A.I. debate can best be summed up as follows: a while back, a joke image of a restaurant server went viral that was generated with an A.I. prompt. The accompanying explanation “highlighted” the red flags that showed it was A.I., circling the background details that “looked off”. Meanwhile, the most-obvious tell, the server having feet for hands, was deliberately ignored. That’s this debate in a nutshell.

It’s okay to be skeptical of A.I. art. I’m skeptical of it constantly! Though it shouldn’t be used as lazy shorthand to brush off genuine mistakes and bad art made by real people. But it is, and way too frequently. We need to be more thoughtful in our criticism, or else we’ll look silly. And that’s bad.

Alternatively, you can ignore my advice and call something you don’t like “A.I.”. The choice is yours.