Black Mirror’s “Joan is Awful” and the plea for quality media in the face of AI

Several people in my life told me I should watch ‘Joan is Awful,’ the first episode in the newest season of Black Mirror. I had never seen an episode of this show, but as an aspiring screenwriter currently working in Web3, apparently it was made for me. All I knew before pressing play was that Salma Hayek was in it (love her) and that the premise involved the creation of an AI-generated TV series. “It’s totally relevant to the writer’s strike!” People kept telling me.

So I pressed play, with very little expectation. If you haven’t seen it, the basic plot is that Joan is a corporate pawn who doesn’t love her fiancé and is flirting with her ex. One night she turns on Netflix (in this world, “Streamberry”), and on her “recommended for you” is a show seemingly all about her, starring Salma Hayek. Streamberry somehow recreated her day in less than 24 hours and projected all of her faults and sins for the world to see. The rest of the episode is about Joan’s attempts to stop the show from airing anymore and get her life back.

About halfway through the episode, we learn how Streamberry does this: they’re mining Joan’s data in real time and using a superfast “quamputer” to generate her life as a TV show, entirely made from CGI. There are no humans behind the production of this show, with the exception of Streamberry execs and the computer whizzes who built the tech. Once it’s revealed that there was literally no humanity involved in the creation of the show, I realized why I wasn’t that interested in this episode: it’s written like an episode of “Joan is Awful.” As in, the humanity just wasn’t there.

Once it’s revealed that there was literally no humanity involved in the creation of the show, I realized why I wasn’t that interested in this episode: it’s written like an episode of “Joan is Awful.”

I don’t know about you, if you’ve seen the episode, but I didn’t care about Joan at all. I didn’t really want her to succeed, I didn’t find her redeemable or even all that interesting. And the plot itself seemed extremely random. It was as if the AI had generated three A-List celebrities to star in a show: Salma, Cate Blanchett (who plays another Joan variant), and Annie Murphy (who plays our reality’s Joan). Even the midpoint, where Joan desperately seeks the real Salma’s attention, felt like asking ChatGPT to create a despicable act for a character to commit in a TV show. Words populating in a chat box: cheerleader costume… defecation in a church… draw a phallic symbol on the lead character’s face. Like… Huh??? Every character felt like a bland archetype, especially our-world Salma whose dialogue was made up of incredibly cringe, inaccurate Spanglish.

None of the characters felt real, because they weren’t supposed to. I believe this was intentional on the writers’ part: it was a warning of the type of “content” we would be getting under this AI screenwriting model. Yes, I found it entertaining, but I wasn’t moved by any of it. And excuse me for still wanting to be moved by popular art and media in 2023.

I think a lot of the commentary and “takes” of this episode that I’ve seen online are forgetting the central point about generative AI that writers and creators are trying to make. Not everything needs to be optimized for efficiency, especially art. And yes, writing is an art. It’s a specialized skill that requires imagination, technique, and, most importantly, a nuanced understanding of what it means to be human. Sorry ChatGPT, but that’s just something you will never be able to generate.

––

Blokhaus is a marketing and communications agency with a focus on Web3 and emerging tech. Since we were founded in 2021, Blokhaus has supported numerous high-profile projects and activations around the globe, working in partnership with some of the biggest brands in the world. To learn more about our work, check out our Case Studies. To get in touch, visit our Contact Us page.